About this Author
College chemistry, 1983
The 2002 Model
After 10 years of blogging. . .
Derek Lowe, an Arkansan by birth, got his BA from Hendrix College and his PhD in organic chemistry from Duke before spending time in Germany on a Humboldt Fellowship on his post-doc. He's worked for several major pharmaceutical companies since 1989 on drug discovery projects against schizophrenia, Alzheimer's, diabetes, osteoporosis and other diseases.
To contact Derek email him directly: email@example.com
November 30, 2005
Reader Jeff G. pointed out this blog post, which quotes the letters column in the most recent issue of Vanity Fair. The magazine ran a profile of Paris Hilton, about whom the less said the better, and quoted her in-the-same-category sister, Nicky, to the effect that she was working so hard, running multimillion-dollar companies and all, and what were the people writing about her doing at her age (21) that was so important?
The magazine published a letter from one Steven Sabes of Minnesota, who said:
. . . I would like to repond to that. When I was 21, I was busy working toward my Ph.D. in organic chemistry at the University of Minnesota. I was the first to synthesize the compound okadaic acid -- shown to be the leading cause of breast cancer.
Well, I'm glad to see some organic chemistry make it into the pages of a magazine as empty-headed as Vanity Fair usually is. Face it, those of you readers who are fellow chem-geeks - our jobs might as well be voodoo hoodoo as far as the VF demographic is concerned. And Sabes and his okadaic acid work appear to be real. He's a graduate of Craig Forsyth's group at UMN.
The problem is, someone at the magazine seems to have mangled his point. Okadaic acid isn't the "leading cause" of breast cancer - it's a marine natural product that few people have ever encountered, unless they've had some really, really bad shellfish. It's a phosphatase inhibitor, which puts it in an interesting class of activity that's been hard to approach through traditional med-chem, and it's been used for many in vitro and in vivo research studies in cancer and other fields.
So I appreciate Sabes taking a whack at the Hilton sisters, and letting people know that organic synthesis actually exists. I just wish that his meaning had stayed intact.
+ TrackBacks (0) | Category: Press Coverage
Despite an AP story today that named a Merck/Banyu research site in Japan, It looks like it's the Merck research site in England (Terlings Park, Harlow, Essex) that will close, as some rumors already had it. A comment was left to the post below that seems to confirm things, and I believe that it's authentic.
I'm sorry to see it. They've been doing good work there for a long time - a large part of the Substance P story that I spoke about here was done at Terlings Park, for example. Here's hoping that Merck doesn't have to cut more in the future.
+ TrackBacks (0) | Category: Current Events
November 29, 2005
There's a peptide hormone called ghrelin that a lot of pharmaceutical companies have worked on in the last few years. It's a good target for obesity (and perhaps diabetes, too), since it's involved in appetite signaling between the stomach and the central nervous system. It's also involved in growth hormone signaling, too, though, so the situation is complicated.
And it just got more so. It turns out, according to a paper in the November 11 issue of Science. A group at Stanford has discovered that the same percursor protein that's carved up to produce ghrelin is also used to produce another peptide hormone that they've called obestatin. That one has its own receptor, and its own signaling network, and it appears to do the exact oppositeof what ghrelin does. Injections of ghrelin stimulate feeding in mice, and injections of obestatin inhibit it, for example. Similarly, ghrelin increases gastric emptying, and obestatin slows it down. (One place where the two peptides don't match up is their effects on growth hormone secretion - obestatin doesn't seem to do anything to the growth hormone axis at all).
So now we know more about the regulation of appetite than we used to, although researchers in that field probably thought it was complicated enough already, thanks very much. What I find particularly interesting about this discovery is how these two opposing hormones are cut from the same larger protein. That means that they both come from the same gene, you know. Which shows you just how far a pure genome-driven approach to drug discovery will get you: not far enough. You'd never know about ghrelin from just reading off human genes, because it's produced after the orginal protein is transcribed. And you'd never know that the same protein is the source for another hormone that negates ghrelin, either. All that complexity is downstream of the DNA. (Update: see the comments for some dissenting voices on this issue).
We already knew that general principle, of course. As soon as the estimates of the total number of human genes starting coming in, it was clear that they were way too low to explain the number of different proteins that we already knew about. But examples like this one just rub it in. . .
+ TrackBacks (0) | Category: Diabetes and Obesity
November 28, 2005
Everyone's been waiting for it, and today the ax finally dropped at Merck. It had to. The company still has over six thousand Vioxx lawsuits piled up, and one of their biggest moneymakers (Zocor, simvastatin) is going off patent (and that's not their only patent problem). Their advanced research pipeline has taken a terrible pounding the last couple of years, too, with the loss of a couple of Phase III compounds and the post-approval death of Pargluva.
It's not easy to tell from the press release, but it looks as if many of the 7,000 jobs that Merck is cutting will come from manufacturing. They're closing five production sites outright and trimming some others over the next two years. Discovery isn't being spared, though, since Merck's also closing a basic research site. (That's how you know things have gotten bad at a big pharma company). No details on which one yet, but I think we can assume that it's not going to be Rahway, and I don't see how it can possibly be West Point, PA either.
That would mean that the folks at Merck-Cambridge and
Merck-La Jolla (Update: whoops - that one's been closed since June) must be pretty jumpy, and I don't blame 'em. I listened to a fair amount of the company's conference call from this morning, and a spokesman said that the employees at each site designated for closure would be notified over the next two days. There won't be any public announcements until then.
As someone who's been through some rounds of closures and layoffs (a memorable one of which happened at almost this exact time of year, come to think of it), my sympathies go out to Merck's employees. I don't believe that their company has ever been through something like this before. I'm sorry to see y'all joining the club. And you people at Pfizer, I'm afraid that your membership will be up for renewal soon. . .
Update: here's a list of Merck's research sites. . .
+ TrackBacks (0) | Category: Current Events
November 27, 2005
One of my colleagues came by the other day to dispute a remark I made in a comment here the other day. The question is the role of "pedigree" in hiring into pharmaceutical research labs, which varies quite a bit between companies. Some are notorious for "old boy" systems - you'll find an awful lot of Corey folks at Pfizer, for example. Does this consideration make sense, and if it does, how much should it count for?
It was pointed out to me that I've consistently claimed that there's no one best model for drug discovery, and that flexibility is a virtue. Thus, went the argument, I shouldn't rule out hiring by pedigree, because there's surely no one best way to select chemists, either. I agree with that last part, but my comment against pedigree was aimed more at people who rule out hiring by any other means. I was actually making a pro-flexibility case.
But that doesn't settle the underlying issue: is hiring based on who the job candidate worked for a good idea, or not? First off, I should note that this is mainly an issue for PhD lab-head jobs. Associate positions get filled by people from all over the place, since the job requirements aren't as broad. With that in mind, I think this can be broken down into categories: the candidate could have worked for
1. Someone with a poor reputation
2. Someone that no one's ever heard of
3. Someone with no particular reputation either way
4. Someone with a reasonably solid reputation, but not a star
5. Someone who's famous
You'd probably stay away from (1), but I frankly can't think of anyone I'd put in that category. A lot of professors in group (2) might be classified that way if anyone paid attention to them, I guess, so that group is at a disadvantage, too.
But they're at a disadvantage for many other good reasons. The main way to stay in that second group is to not publish much, not give talks at meetings, and work at a school that's not known for chemistry. Clearly, you don't want to come out of a research group in this category if you can help it.
Category (3) is a step up, but not by much. Professors in this group are not totally unrecognizable, but they're not known for anything specific. You recall hearing their name, or seeing them in a journal, but you didn't read the paper. Closer inspection might reveal that to be either their fault or yours - it could go either way. The reason you don't have a handle on the person might just be that no ones finds their work interesting or useful.
Category (4) is where the arguing really gets going. There are an awful lot of research groups in this category, and the quality of the people they turn out varies widely. My colleague argues that picking from category (5) instead gives you a bozo filter, that you're at least assured of some level of quality control.
My reply was that if someone is a bozo, you should be able to pick up on that during the interview process. Personally, if I had to hire a hundred people from only one category to staff a med-chem department, and boy am I glad that I don't, I'd go with (4) over (5) as long as I got to select from the pool. If I had to hire blind, though, category (5) people might well give a higher average.
But it could come at a cost. Just as there are all sorts of people working for not-so-famous professors, there are all sorts of famous professors. You've got your up-and-coming stars, your over-the-hills, your genuine geniuses and your sweatshot slave-drivers. Some of these folks damage as many students as they improve. (And there's another downside to hiring only from the elite: some of them come complete with bonus attitude).
+ TrackBacks (0) | Category: How To Get a Pharma Job
November 25, 2005
Well, like many of my fellow Americans, I'm taking today as the second day of Thanksgiving. The Wonder Drug Factory where I work is closed for the day, and I'm here at home with my two children while my wife braves the shopping crowds. (I'm positive that I got the better end of that deal). My hardest task will be keeping the two of them from eating the rest of the chocolate pecan pie I made, which they've already decided would be a reasonable breakfast. And no, I have no reason to complain about anything if these are the sorts of problems that occupy me: Thanksgiving, indeed.
At any rate, I can recommend a browse through the blogroll over there on the left. There are some new additions today, namely TP With Page Numbers, Aetiology, Adventures in Ethics and Science, Politics and Ethics of Science, and the fine new group blog Nature Erratum. (That last one makes me wonder why no political blogger has titled a site "The New York Times Regrets the Error"). And if you don't know the medblogger Dr. Charles, then you've got even more to read instead of doing work. I'll see everyone on Monday!
+ TrackBacks (0) | Category: Blog Housekeeping
November 22, 2005
If you stopped one hundred people on the street and asked them to name a drug company, I'd be astonished if a single one of them mentioned Serono. But they're one of the largest biotechs in the world, even though their profile is low. Being a privately held (indeed, family-owned) concern surely has something to do with that, because if people can't talk about your stock, they often don't talk about your company.
Serono has been around for about a century, and for a long time they made their living in hormones and fertility treatments (some of which were extracted directly from urine, which must have been a joyful task). They really took off though, in the last ten years, making the current family CEO Ernesto Bertarelli one of the hundred wealthiest people alive. Their recent growth has been due to Rebif, a recombinant beta-interferon for multiple sclerosis, but before that one of their big products was Serostim.
And that's one part of the company history they'd like to forget. Serostim is a growth hormone preparation approved in 1996 for treatment of HIV-related wasting. But newer antiretroviral drugs came on the market very soon after that and AIDS wasting became less of an issue (in fact, some of the HIV protease inhibitors are well known for redistributing and perhaps even adding body fat). Serono fought back, as any company would, but they comprehensively crossed the line.
Their first tactic was to promote a medical device to measure wasting in HIV patients, which gizmo (wouldn't you know) indicated that people needed Serostim even though they looked fine. The idea was, er, that their cells were losing mass, even though their outward appearance might not indicate it. Another campaign tried to promote the same device (and Serostim) to diagnose and treat the adipose effects of the retroviral drugs. Neither were approved for such a use, as you might well have guessed. And their third method was more to the point: to flat-out pay physicians for the number of Serostim prescriptions they wrote, in one notorious case by picking up the tab for a free vacation in the south of France.
The company recently settled with the US government, agreeing to their guilt and paying $725 million in fines. It's worth noting that the whole scheme was done in by five employees with knowledge of the matter, who will now share some $50 million of the fine under Federal whistleblower statutes. (This seems to be a perfect example of what the law was designed to do).
So having put this behind them, Serono finds themselves in the position of several other companies over the years, with most of their revenue coming from a single product and not much else looking fit to replace it. And, as some other companies have done in such times, they've called in the investment bankers.
But I'm not sure who's going to line up to buy them. Right now, you'd be getting Serono for nearly the highest price it's ever commanded. If someone decides that they want Rebif for the rest of its patent life, that would be the best reason I could think of to go ahead with a deal. . .of course, it would have to be someone with a huge marketing arm and the conviction that they could make more money than Serono could with the stuff, which I'd guess narrows it down even more to companies whose names start with a silent "P".
+ TrackBacks (0) | Category: Drug Industry History | The Dark Side
November 21, 2005
This Volokh post (via Instapundit) about a former gang member who's been "nominated for a Nobel Prize" prompted me to leave a comment there, which I'll expand over here. It would seem that many people don't realize where Nobel Prizes come from.
The Peace and Literature prizes have a comparatively open nomination process, which makes for what I'm sure is a pretty poor signal/noise for whoever handles their mail. (Of course, the signal/noise of the list of eventual winners for those two isn't so great, either). But the science prizes are run in a tighter fashion. Here's the nominating committee for the Physiology/Medicine prize, for example, and it's very similar for the Physics and Chemistry Prizes.
The various Scandinavian professors involved are notoriously quiet about their choices, as are most prior laureates. The committees never say who these "other scientists from whom the Academy may see fit to invite proposals" might be, and I'm sure that identifying oneself would be a sure way to be dropped from the list. That's not to say that there are no controversies, just that we don't get to hear about them in detail for fifty years or so. That link will let you search older Medcine prizes. It's interesting that the corresponding database searches for Physics and Chemistry aren't even available.
What this means in practice is that no scientist, in theory, is able to be identified as a "Nobel Prize nominee." That doesn't keep it from happening, though. In fact, that link will take you to the story of someone who is claimed to have been nominated five times. A Google search for "five-time Nobel Prize nominee" turns the same person up all over the place. It's almost as if he hasn't done anything to discourage the practice.
One of most notorious recent examples was a neurologist, William Hammesfahr, who was all over the media during the Terri Schiavo case. He was invariably referred to by his supporters as a Nobel prize nominee, but this was another whopper. At least he only claimed to be nominated one time, but anyone who claims to be nominated at all should be under suspicion.
Searching for "N-time Nobel Prize nominee" for various values of N will net you all kinds of stuff. Excluding the Literature and Peace candidates, you find Nigerian crank physicists (more here, medical quacks, and Dr. Johanna Budwig herself, the current record holder in this doofus category. She's usually described by her acolytes as a "seven-time Nobel Prize nominee", which wouldn't be good news even if it were true, wouldn't you think? I note with amusement that in she was being called merely a six-time nominee back in 2002. Things have clearly advanced since those days, which is remarkable since I don't believe that Dr. Budwig is actually still with us.
+ TrackBacks (0) | Category: Snake Oil
I wanted to let everyone know that I have an article up on Medical Progress Today, wondering where all those safe drugs we grew up with have gone. . .you know, aspirin, acetominophen, penicillin, that sort of thing. . .
+ TrackBacks (0) | Category: Toxicology
November 20, 2005
I've written before about the gene known as SIR2. Overexpression of it (or its homologs in different animals) extends lifespan in a range of organisms, and there's been a tremendous amount of research on these over the last few years. A good deal of evidence has linked them to the known life-extending effects of caloric restriction. In mice, for example, Sirt1 is involved in nutrient sensing and fat mobilization.
It suggests a pretty neat package, but the ribbon on it is unraveling. I wrote about Sir2/Sirt1 here a couple of years ago, where I said "An extra copy of the gene lengthens life; deletion shortens it." Well, in yeast cells that appears to be true, when you measure lifespan by how many times the cells can divide before burning out.
But what if you measure lifespan by the amount of time the cells can live when they're not dividing? That's the subject of a new paper in Cell from Valter Longo's group at USC, which they have given the provocative title "Sir2 Blocks Extreme Life-Span Extension." Yep, deleting it actually extends the non-dividing lifespan of the yeast, and combining that with caloric restriction increases it even more. These yeast cells have some problems, though, some of which can be ameliorated by further mutations in the IGF-1 pathway (itself heavily implicated in metabolic rate and lifespan). Yeast with combined Sir2 and IGF mutations, under caloric restriction, live longer by up to sixfold, a startling increase.
So what about higher organisms? Well, there appear to be some very similar findings in mice - maybe. Earlier this year, Frederick Alt's lab at Children's Hospital in Boston deleted Sirt1 in mouse cells and found, quite to their surprise, that the cells were extremely vigorous indeed. Such cell lines start to break down after a certain number of passages (cell divisions), but the Sirt1 knockouts just keep rolling along.
They then tried growing the cells under oxidative stress, but they plowed right through that, too. That led to the thought that Sirt1 might be some sort of checkpoint, which would normally limit cell division under such DNA-damaging conditions. But the Sirt1-deleted cells showed no signs of greater DNA degradation than normal lines. They're quite robust.
This is all extremely interesting, but you may have noticed that I pulled a fast one here. In these mouse cell lines, it appears that replicative life span has increased when Sirt1 is taken out - but with Sir2 deletion in yeast that's not the case at all. There it's replicative life span that takes the hit, and non-replicating (chronological) life span that's increased. How do we reconcile these blatantly contradictory findings? I've no idea, but it's a safe bet that several high-powered labs are currently working overnight shifts to answer that question.
So much for mouse cells - how about whole mice? Well, as hardy as some of their cells may be, the Sirt1 knockout is a pretty hard animal to prepare, because most of the mice don't survive. The ones that do appear fairly normal, but have a complex phenotype, which includes decreased fat mass and body weight. (Alt's group is also trying to interfere with Sirt1 in adult animals, bypassing all the developmental roles that make the standard knockouts so hard to work with).
The big question now, given all these divergent cell findings, is: will these guys live longer, or not? And what happens to them when you put them on a limited-calorie diet? Are they going to act like the replicative-aging models, or the chronological aging ones? (We'll leave the yeast-mouse contradiction out of it for a while). Perhaps the two mechanisms will fight each other to a standstill, leaving the animals with plain ol' normal lifespans, but with some tissues acting much younger than the whole-body age and some acting much older. Mice generally live around two years. I wonder just how many months ago these lifespan studies started. . .
+ TrackBacks (0) | Category: Aging and Lifespan
November 17, 2005
So, what's Kevin Trudeau up to these days? (Besides counting those great big stacks of money, that is). Back in August he was selling huge loads of his "Natural Cures" book, which is a rather huge load in and of itself, even though people seemed to be catching on to the fact that there are no cures in it. The New York State Consumer Protection Board certainly caught on back then, calling Trudeau's book a fraud "from cover to cover".
Well, he's still moving truckloads of the book, much to my disgust. Demolition jobs like this one in the major media don't seem to have slowed things down much at all. Still, Trudeau seems to have branched out a bit. Customers are now complaining that they're being signed up for his newsletter ($71, don't you know) when they order the book by phone. Oddly enough, it proves to be difficult to cancel the newsletter (and its $14.95/month charges), since no one seems able to get through to customer service. No doubt the next phase will see people involuntarily signed up for a "lifetime membership", which runs a mere $499.
Not content with that, the latest report is that he's reselling his mailing list - which must be sizeable - for even more money. And as you'd expect, he only deals with the classiest marketers:
'Customers who call one of Trudeau's toll-free order numbers have received spooky junk mail from a Nevada company, claiming to be a 'secret society' with 'personal knowledge' about these consumers. This junk mailer calls itself the 'Nouveau Tech Society,' but it is actually another book publisher in Henderson, Nevada that sells a $150 book teaching people how to rule the world. . ."
This appears to be the Neo-Tech/Zonpower people, who are a batch of crazed scamsters from way back. I thought about putting a link in there, but the tiniest Google search on any combination of those names immediately throws up dozens of pages of over-the-cliff weirdness. (You'll go from a standing start to UFOs full of Bavarian Illuminati very quickly, so be warned.) Quite a crew to work with, I have to say - the only thing that I can be sure if is that if they're dealing with Kevin Trudeau, their checks must have cleared.
Trudeau's response to this complaints? A press release of breathtaking arrogance, titled "Bureaucrats Try to Interfere With the Sale of Trudeau's Best-Seller." (That link is worth reading, although if you have a tendency to high blood pressure, I'd think twice.) As usual, Trudeau wraps himself in the First Amendment, pretending to having never heard of the concept of commercial speech.
So, Mr. Trudeau, just for you I'm going to exercise my First Amendment rights. In my considered opinion, you are a fraud. A con man. A shameless opportunist and peddler of lies and gibberish. You are urging cancer patients, HIV patients, diabetics and people with every other life-threatening disease under the blue sky to throw away their drugs, stop seeing their doctors, and send their money to you in exchange for all the dust you can shake from your expensive shoes. Enjoy your money, won't you?
+ TrackBacks (0) | Category:
November 16, 2005
I had some email from an undergraduate chemistry major who's interested in doing drug discovery work eventually. He was wondering if organic synthesis was still the way to go in graduate school, and my answer is "definitely." I've warned people away from too much of a medicinal chemistry focus in their graduate work before, and I'll be glad to do it again.
Let me be clear, though, that I'm talking about academic medicinal chemistry programs as they've usually been run. What you often end up doing in these is learning a fair amount about each of the major areas of drug discovery, but you generally don't get really good at any of them. And since there's no market for a one-man drug company, you're left in a bad situation. We don't need fair-to-middling chemists who are also fair-to-middling pharmacologists. Perhaps there are other ways to run a med-chem department at a university that won't lead to this problem, but the only one I can think of is to run it like a little drug company, which I don't think is going to work out very well.
The reason we like for people to do lots of organic chemistry before they join the drug industry is that we (the medicinal chemists) are the only ones who understand that stuff. Other departments have members (sometimes) with a reasonable knowledge of the main points of organic synthesis, but it's only the chemists who can really dive into it. Someone has to, and by golly, it's us. So we need people who really know what they're doing: people who can use the fastest and cleanest routes to making the most diverse analogs, who can think up structures that no one else has ever made and reduce them to practice, and who can find the cheapest, most reliable ways to scale up a synthesis for real world use. The wider the range of serious organic synthesis experience you have, the more we'd like to talk to you.
This might eventually become a problem for us in industry, as the field of organic synthesis continues to mature. Already, I think we have a bit less emphasis on people who've done total synthesis of natural products, because fewer groups are doing that these days. (And some of the groups who are, aren't doing it in a way that everyone in the group gets the kind of training we need - the "team of lab Sherpas" approach damages as many people as it improves, as far as I'm concerned).
But total synthesis, done right, is still the perfect sort of training for our needs, even though we don't do thirty-two step reaction sequences. It sends you all over the state of the synthetic art, gives you a varied problem-solving workout, and trains you to always look for alternatives to the chemistry you're doing. We couldn't ask for more. I find this state of affairs a bit irritating, though, since I think that total synthesis is a slowly dying art form, and rightfully so. There's less and less need, in my opinion, for those thirty-two step routes. I can almost imagine keeping it all going just to provide the drug industry with the kind of people it likes to hire, though. In fact, I can almost imagine that this is already happening right in front of us. . .
One of the major alternatives to total synthesis is working on new synthetic methods, which is a valuable service to the science. But while that can train people very well for drug research, it can also leave them partially crippled. It depends on how broad their experience of different reactions has been, and how general their approach to problem solving has had to be. If you come out of graduate school as the world's best set of hands for one particular reaction, my advice is to go do a completely different sort of post-doc, to prove that you have the hands to do something else. And that's because "do something else" is pretty much the job description of a working medicinal chemist.
+ TrackBacks (0) | Category: How To Get a Pharma Job
November 15, 2005
The American Heart Association meeting has been going on this week, and it's one of the high-profile venues for cardiovascular research results. Today, among other things, came the details of a head-to-head trial (called, in standard cute-acronym style, IDEAL) of Pfizer's Lipitor (atorvastatin) against Merck's soon-to-be-generic Zocor (simvastatin). Pfizer's worried about that change in status, since this will be the first of their major statin competition to go off patent - the fear is that HMOs will insist on a switch to the cheaper medicine as quickly as possible.
So an 8,800-patient trial was run, using patients who had already had one heart attack (although not necessarily a recent one). One group got high-dose Lipitor, and the other got the standard dose of Zocor, and they were followed for an average of nearly five years. This was one expensive trial. The hope was the the Lipitor regimen, which lowers LDL more, would show a beneficial effect. Which it did - and which it didn't.
Pfizer's problem is that they missed their biggest target, a reduction in "major coronary events". The Lipitor and Zocor groups were indistinguishable. In some secondary categories, though, Lipitor came out ahead, such as when you look only at the frequency of nonfatal heart attacks. But overall cardiovascular mortality was the same for both groups. So Pfizer can claim some benefit from Lipitor, but not in the categories that would immediately convince cardiologists. I'm sure they'll be giving it the good ol' Groton try, though
All this is quite interesting, since Lipitor had already vanquished another statin, Pravachol, in a study in recent heart-attack survivors. Merck, meanwhile, had also tested Zocor in angioplasty patients, with disappointing results as compared to Lipitor's effects in a similar patient population. You can bet that Pfizer expected more from this study, and I'm still a bit puzzled that they didn't get it.
Analysts have already cut their sales estimates for Lipitor next year, figuring - surely correctly - that these results will cause some shift toward generic simvastatin. That doesn't do Merck much direct good, but it leaves them with an opening for their combination therapy Vytorin, which is Zocor plus Schering-Plough's cholestrol absorption inhibitor Zetia. And as fate would have it, they're running a big comparison of Vytorin to Zocor right now, since the darn stuff is going off patent, anyway.
The earlier comparison trials had already poked a hole in the "lower LDL is always better" hypothesis, which has been one of Pfizer's arguments for Lipitor. (It's also the first place that Merck and Schering-Plough chose to attack them). Still, they always had good clinical data to point at, too, but now that advantage has been eroded. So in the last fifteen months, we've had the spectacle of Merck trying to shoot down Lipitor and winging its own drug, and then Pfizer aiming at Zocor and blowing a hole in Lipitor instead. What a business. Can anyone name me another where things like this are a way of life?
+ TrackBacks (0) | Category:
November 14, 2005
Andrew Stimpson isn't a scientist. If he were, he might have heard the line about extraordinary claims requiring extraordinary evidence. And his claim is indeed extraordinary - he says that he has managed to clear HIV from his system without therapy. Well, other than vitamins, don't you know - and if he says that he got them from Matthias Rath, I'm going to have to go lie down for a while.
This story has generated all sorts of irresponsibly breathless headlines, but when you read the stories underneath them you find that there's not much behind it. Stimpson received a positive HIV diagnosis on the basis of two antibody tests in 2002. Near the end of 2003, he was found to be negative, and was asked to repeat the test to confirm it. He refused, and sued the British health trust that did the testing.
Update: Press reports disagree about this. Some have the story as above, and others say that Stimpson tested negative on several occasions during 2002 and 2003. His initial postive tests showed what is being described as "an extremely low viral load."
Here's where things get messy. Stimpson appears to have sued because he felt that the original tests were in error. (The agency naturally stood by both its positive results). When no money was forthcoming, he then seems to have gone to a couple of the British tabloids with his miracle recovery story: ". . .I am just one person who managed to control (HIV), to survive from it and to get rid of it from my body", he's quoted as saying, which is an interesting statement from someone who was previously claiming not to be infected at all. Update: Stimpson eventually received a letter from the National Health Service calling his HIV-negative status "exceptional and medically remarkable", so he at least didn't come by his miracle-recovery story alone.
Stimpson hasn't been tested again, and doesn't seem to be available at the moment. I am not inclined to believe any claim such as his, to put it mildly, until he's been poked and prodded from every angle - to put it mildly. You would think that he might wish to help other HIV sufferers if he really has reversed the disease, wouldn't you? The article link above quotes the head of a charity in England as saying "The answer may turn out to be very complex. We must not jump to conclusions." Actually, I'm close to jumping to the conclusion that the answer might be rather simple.
Update: Press reports also disagree - markedly - about Stimpson's willingness to undergo further tests. My final sentence isn't meant to suggest some complex biochemical rationale. I'm thinking that the chances are best that the first positive results were in error - after all, their false-positive rate is surely much higher than the spontaneous-clearance-of-HIV rate. The health agency that tested him, though, might well prefer to treat this as an amazing medical anomaly rather than as a botched test. And given that he wasn't able to collect damages, it became in Andrew Stimpson's financial interest to go with that explanation as well, selling his story to the British tabloids for an undisclosed amount.
In the end, I agree with this quote, from the Nature.com news item on this story: ""If it is real, it's very interesting," says Jonathan Weber, an expert on infectious diseases at Imperial College London. But he cautions that the most likely scenario based on the current evidence is "either a false positive [in 2002], or he's still infected".
+ TrackBacks (2) | Category: Infectious Diseases
November 13, 2005
I haven't worked in a US academic chemistry lab since 1988, so you'll have to take that into account as you read today's post. But I don't think that things have changed enough to invalidate this observation: many grad-school science labs are so depressing as to defy belief. This isn't universal, but I've seen enough examples to convince me. The atmosphere doesn't correlate well with the amount of money around, either, because I've seen some lower-level departments that weren't so bad, and a couple of Ivy League lab corridors that would pull the serotonin right out of your brain just to walk down them.
Many of the students and post-docs working at these places don't realize this, though, which is surely to their benefit. It's only after you've gone out into the Real World for a while and come back for a visit that it hits you. That's certainly how it dawned on me. I believe it says over there to the left that I went to Duke: maybe someone there could tell me if that second-floor "graduate student lounge" in the chemistry building is still there? I sure hope not.
This was one of the most cheerless rooms I've ever seen; it made laundromats seem warm and inviting. There was no ceiling as such, just the fluorescent lights hanging down from the industrial clutter above. Some scuffed paneling on the walls surrounded an assortment of worn, stained, mismatched thrift-store furniture. A small damaged table in the middle of a couple of sprung couches held an assortment of torn, dogeared cycling magazines, some of which had been there when I arrived and were still sitting there when I defended my PhD. A scarred counter held a 1970s-vintage microwave, which might as well have had "X-1 Prototype" stenciled on it. A bike frame without wheels, furry with dust, was chained to a rack in one corner. That was still there when I left, too.
You feel bad enough at 3 AM on a Sunday morning up in the lab, running a reaction for the twenty-third time. Taking a break by wandering down to a dingy room full of junk is not the recommended antidote. Why more people didn't just decide to end it all after a session in there is a real mystery.
Update: Reader LNT, in the comments to this post, passes on this news:
"I graduated from Duke in 2001 and the graduate student lounge hadn't changed from what you described. I don't know why university chemistry departments can't budget a couple thousand bucks a year to keep a decent communal area for their slave labor pool..."
+ TrackBacks (0) | Category: Graduate School
November 11, 2005
I mentioned that the Wall Street Journal is running a free-access week. That'll give you a chance to check out the Nov. 11th column by their science columnist, Sharon Begley, who sounds an appropriate note of caution about genomic-based drug discovery. (And I'm not saying that just because she quotes me). It'll be good to get these concepts out to a wider audience.
+ TrackBacks (0) | Category: Press Coverage
November 10, 2005
You know, after all the philosophical wrangling that's gone around here the last couple of days (I refer to those record-setting comment threads below), I have to say that there's something about scientific research that I really appreciate: things get resolved.
Not everything gets resolved, true, which is also part of the fun. But enough things do get settled to provide a person with a sense of accomplishment. In medicinal chemistry, we have to come to some firm conclusions about things, for example: Is Drug Candidate X more efficacious than Drug Candidate Y? How long does it last in the blood after an oral dose? Is it more toxic? What will it cost to make?
Naturally, there's room to argue about the details of all those things. Try "efficacious", for example - in which model of the disease are we talking efficacy? Efficacious by which biological criterion? Are those both the right ones to use to try to predict clinical success, or are we just kidding ourselves? (A constant temptation, that). The other questions can be exfoliated in the same way. Which species are we measuring blood half-life in? Is that the relevant one? How long do you think the half-life should be, and why do you think so?
And you'll have to define "toxic" in the same manner that you had to define "efficacious". Toxic at what level, in what way, and in what species? Is that result relevant to man, or just another stupid distraction? How do you know, and just how much are you willing to bet on that opinion? A million dollars for multiweek tox studies? Tens of millions to get started in clinical trials? Hundreds of millions to get the thing to market? The whole company if you're wrong even after that?
No, there's enough uncertainty to keep things lively, all right. But there are still a lot of things that get settled along the way, once and for all. This reaction really is more reliable than that one. That chiral methyl group really is pointing in that direction. This compound really does bind more tightly to the target than that one and no, we really, really aren't going to develop that other one that just killed off all the rats. After a week of philosophical tug-of-war (for which I have only myself to blame; no one forced me to write about Intelligent Design), I do enjoy the certainties of a clean NMR spectrum
+ TrackBacks (0) | Category: Life in the Drug Labs
November 9, 2005
While the Wall Street Journal is opening its site for free this week, may I recommend this excellent article on the vaccine and antibiotic markets? It's a clear-eyed look at why drug companies haven't put more time and money into these areas over the years. The headline makes it sound as if it's going to be a pharma-bashing-festival, but the authors (Scott Hensley and Bernard Wysocki) lay out the facts, which are just as I understand them from my vantage point, too.
The second article in the series is also up here. It's an equally good overview of the possible incentives that are being discussed to encourage work in vaccines and anti-infectives. I'm glad to see the idea of incentives being discussed, because as it stands, the market isn't necessarily going to give us what we need in the time we need it. New antibiotics are generally reserved for use in resistant cases only, so you can't make your money back there. And new vaccines can end up costing too much in liability suits (many - most - of which aren't particularly well justified). But put some incentives in there, and perhaps the numbers can work out. The article goes into detail on some of the proposals - straight cash, guaranteed purchases, extra product exclusivity, and so on.
I know that some people will hear these ideas and wonder why the government doesn't just do the research itself, rather than cough up money to the drug companies. The biggest reason is that the drug companies are better at it, and faster as well. We stay on our toes competing against each other. The biggest pitfall in these incentive plans, as far as I'm concerned, is that it might end up with companies that have no one breathing down their neck. Better to have two or three organizations racing each other and throwing elbows to grab the prize, than to have someone ambling over to pick it up.
+ TrackBacks (0) | Category: Infectious Diseases
November 8, 2005
There's an interesting letter to Science in the latest issue (Nov. 4, #5749, p. 777), in response to their special section on drug discovery in the July 29th issue. Adrian Ivinson, a former editor of Nature Medicine and now head of a new research center at Harvard Medical School, writes that the section:
". . .did not recognize an increasingly relevant but underappreciated and underutilized role for academic research in drug discovery.
Universities invest may millions in basic research that exposes disease mechanisms and therefore unearths new targets. Yet few have invested in the relatively modest infrastructure required to put their discoveries to the test. As a result, many promising targets gather dust on the university shelf. . ."
Really? Send 'em over here. I've spent a lot of time defending the way the drug industry takes basic research from academia and turns it into applications. (See the September 9th, 2004 post here and work up from there if you're interested). The usual complaint is that that's all we ever do, so it's refreshing, in a weird way, to hear a complaint that we're not taking enough. But if these targets are being published somewhere even semi-reputable, believe me, we're seeing them.
And as for the "relatively modest infrastructure", that depends on what you mean by modest. For example, the research site I work at does no manufacturing, no human trials, no large animal toxicology studies, and very little scale-up chemistry (just enough to get through two-week rodent runs). But we have hundreds of people working here, in several rather large and expensive buildings crammed full of expensive stuff. Now, it's true that we're working on a number of projects simultaneously - just how many, I'm most certainly not going to say. But you'd need a lot of this stuff around no matter how few projects you were developing.
Dr. Ivinson goes on to say that assay development and validation, compound screening, medicinal chemistry and preliminary animal tests are functions "well suited" to academia. Perhaps, perhaps. But it should be noted that there are some well-known people (such as Stuart Schreiber) with experience in both academic and industrial research who worry about academia's ability to do this sort of thing. He also says:
"Demonstrating a credible mechanism and target, proprietary lead compounds, and preliminary in vivo efficacy will be enough to bring some of our industry colleagues back to the table."
That it will! Be prepared, though, to drop more than a good-sized grant application's worth of money to do that, though. It's harder than it looks to get that far. And those proprietary compounds might scare away as many companies as they attract, by the way. Proprietary means, of course, that you guys own them, and that means that we have to buy them. We'd naturally much rather have our own compounds. That would mean demonstrating proof of concept with something that's not patentable, but there are worse things. We can always screen, and believe me, we have a lot more things in our screening files than you do.
As I've said, I think that Dr. Ivinson is underestimating the difficulty of drug discovery, but at least he realizes that it's worth doing. The letter finishes with a sentiment that I can only applaud:
"But this will only happen when academics stop treating drug discovery as the intellectually inferior domain of the commercial sector and start seeing it as the natural development of their research."
+ TrackBacks (0) | Category: Academia (vs. Industry)
November 7, 2005
[Update: reading this post, I can see that I was in a pretty testy mood when I wrote it last night. Intelligent Design does that to me. So if you're not in the mood to be ranted at, come on back tomorrow and I'll see what I can do for you. . .]
Further update: comments have now been turned off, to keep this one from rising from the grave. No doubt I'll post on ID again eventually, so everyone will have another opportunity to ventilate their opinions.
OK, one more on this topic, and then we'll try to give it a rest until the Dover school board decision comes down. (The comments to the yesterday's post are still rolling right along, though, as you'd expect from a debating ground like this one). The article by Jerry Coyne I linked to yesterday gives some good anatomical arguments against intelligent design. But I wanted to zoom down to the molecular level for a minute, since after all, I am a chemist.
DNA is a wonderful molecule, no doubt about it. And to someone like me, who believes that the evidence for evolution is overwhelming, it's also a fine illustration of how it works on a molecular level. Others, though, no doubt see in its intricacies the hand of a creator. What, I wonder, are we then to make of the degraded remnants of old viral DNA in our genome? Designed in there, or not? Or what about the long stretches that seem to do nothing but repeat the same few base-pair letters over and over - dozens, hundreds, or thousands of times? Doubtless the Designer would have his reasons, but perhaps some of these would have been better implemented with repeats that aren't so prone to breakage and mismatch. Hundreds of terrible diseases result. (That page is only the barest sample. It's an awful topic to research). It's almost as if these things persist as the residue of ancient random choices or something.
Moving on to what are supposed to be the normal genes, we find entire books can be written on the horrible consequences of tiny changes in the genetic code. Take the so-called Swedish and Dutch mutations in the amyloid precursor protein. Switch the DNA a bit, and you get a new amino acid in the protein. Get the wrong one, and you die, most terribly, from early and rampaging Alzheimer's disease with complications. Those particular mutations have been in families for hundreds of years now - we've tracked them through the generations. They're still with us because the people involved live long enough to have children - many of whom are destined to die the same terrible way - before the underlying disease finishes them off. It's almost as if the consequences of a mutation were more severe when it affects reproductive fitness.
Mysterious ways, mysterious ways. No doubt that accounts for why we (and guinea pigs, and Peruvian fruit bats) can't make our own vitamin C, the way the other mammals can. Or why our livers respond to the excess of free fatty acids in type II diabetes by. . .making more sugar, which is exactly what the body doesn't need. There must surely be a reason, too, a good well-designed one, for autoimmune diseases: having our bodies tear themselves to pieces on a cellular level; I can't wait to hear why that feature was built in. It's almost as if once we've had children, just about anything can happen to us.
I'll stop there. I could go on for pages. Suffice it to say that when I look at the biochemistry of living systems, I see an amazingly complex system, wonderful to behold. And it's held together with duct tape, chewing gum, and weathered pieces of wood - whatever was handy, and whatever worked. It's almost as if it's just been tinkering along for a billion years.
+ TrackBacks (2) | Category: Current Events | Intelligent Design
November 6, 2005
Friday was the end of arguments in the Kitzmiller v. Dover Area School District trial on the teaching of "intelligent design" in the local 9th-grade biology class. We won't see a decision in the case for a while (perhaps by the end of the year), and no one knows how broadly the judge in the case might be inclined to rule.
I don't see how there could be much uncertainty in my position on this matter, but just in case: I think that "intelligent design" is pernicious nonsense. I understand why some people believe it, but the argument from incredulity doesn't do much for me. If I threw up my hands at everything that seemed to complicated for me to explain, I'd be out of a job, and rightfully so. My scientific predecessors kept trying to explain mysteries - good for them! - and I'm not going to stop looking for answers, either.
Since the organization defending the ID position has said that they want to "use the courts to change the culture", here's hoping that they get an enormous bucket of cold water poured on them. I was a college student in Arkansas when Judge Overton ruled in McLean v. Arkansas, an attempt to mandate the teaching of "creation science", and his opinion still makes fine reading. It put the brakes on that whole approach to ridding curricula of evolution, but eventually such selection pressure led to the spread of this latest mutation. "Intelligent Design" is clearly the scion of "creation science" - try as I might, I don't see how anyone but a fool can believe otherwise. If it too gets struck down, we can all expect yet another variation in another few years as the anti-evolution forces continue to evolve.
+ TrackBacks (0) | Category: Current Events | Intelligent Design
November 3, 2005
Everyone will have heard the news that Merck won their second Vioxx court case in New Jersey this morning. This came as a relief to me and to people like me, for several reasons. The immediate one was just that Merck had survived this latest round, of course. This wasn't a particularly strong case, and I had hopes that Merck would prevail, but you never know how juries are going to run.
And that brings up the second reason: if Merck had lost this one, they could expect to lose plenty more. The plaintiff in this case survived his heart attack, had a history of heart trouble and had other risk factors, and doesn't seem to have been a diligent user of Vioxx at all, which he only took for two months. I'm sure that hundreds (thousands?) of other cases could be found that rise to roughly this level, and Merck would likely be crippled by losing them. Now the focus shifts to a Federal court case in Houston, which starts later this month. I hate to put it in these terms, but this next one is going to be something of a tiebreaker.
I'm not saying that Merck should necessarily win every case, though. Vioxx does seem to carry some cardiovascular risk, Merck does seem to have charged ahead with it, and so many people took it that there must have been some people injured. But the FDA did approve the drug, let's not forget, and it's quite possible to argue that its benefits still outweigh its risks. And even if Merck were to win every trial from here on, they'e still be out a huge amount in legal fees. They've taken a beating, both in their reputation and in their finances, and that's not going away.
So no, I don't think Merck should be able to suddenly make all their troubles go away (not that that's going to happen). But neither do I think that they should be driven into the ground like a tent peg by repeated legal hammer-blows. Drug companies should be punished when they screw up. But destroying them for it, in a chancy industry like this one, will just ensure that we don't have many working drug companies.
+ TrackBacks (1) | Category: Current Events
November 2, 2005
In the first post below, I said that we would be in big trouble if we had to rely on a vaccine for all of our protection against a flu pandemic. The problem is, if we're relying on antiviral drugs, we're in even worse shape. (See my post on this from last March.)
Right now, the only drugs likely to do much of anything against an influenza pandemic are the neuraminidase inhibitors like Tamiflu and Relenza. But the problem is, these drugs really have to be taken early in the course of the infection to be most effective. By the time many people realize that they have the flu, it may be too late to do much about it.
One way to get around that problem would be to take the drugs prophylactically, but that has two serious disadvantages. For one thing, this route might well lead to a quicker development of resistant viral strains, which is something that we already know can happen. And for another, it would burn through huge amounts of drug, and we don't happen to have huge amounts of either one.
Why is that? Well, for one thing, neither compound was selling very well until recently. The companies involved have never had to ramp up production to the levels that people are talking about now. It's doable, but it won't be fun. The ten-step Roche synthetic route to Tamiflu uses some azide chemistry, which is potentially toxic and explosive, but it's nothing that a good bunch of industrial chemists can't handle. (It helps, for example, that India's Cipla already has experience making AZT, because that relies on similar chemistry). But any ten-step route is not going to be trivial to implement if you've never done it before, azide or no azide.
A bigger problem is that these drugs have syntheses from a starting material called shikimic acid. That's a component of an important metabolic pathway in plants. (It's important enough that the well-known herbicide Roundup works by shutting it down). Shikimic acid is found in small quantities in a lot of plant species, but star anise, a spice used in Chinese cooking, has a lot of it. (If you'd like to extract some from any star anise you have in your kitchen, here's how). Roche already has a network of suppliers in China, and the generic companies who plan to produce the drug are having a hard time sourcing the shikimate. It can also be produced by fermentation, which Roche uses for some of its supply, but that's an even more specialized process.
All in all, I think it's prudent to stockpile these drugs, although I'm not sure, for the reasons given above, where the US government is going to find the quantities it's looking for. But even if we can pile the stuff up to the rafters, we have to be ready for the possibility that these drugs may or may not do us much good. I see that I've ended both of these posts on the same note. . .
+ TrackBacks (1) | Category: Infectious Diseases
The government's proposed plan for dealing with a flu pandemic is worth some comment, although it's going to undergo mutations just as surely as the viruses do. It's hard to argue with the overall approach, but there are some details that need explaining.
For example, there's a proposed boost for research into vaccine production through cell culture techniques (as opposed to the famous chicken-egg methods), and I think that this is a fine idea. Unfortunately, it was just as fine an idea two or three years ago, and we'd probably be in better shape now if this idea had been pushed back then. Money doesn't convert to time quite as easily in basic research as it does in some other areas (although it doesn't hurt, true). Some of the companies that do work in this area are pointing this out today, in rather testy tones of voice:
. . .so far the government has not backed development of three cell-based vaccines that have received or are close to receiving regulatory approval in the United States and Europe - including one developed by a Meriden (CT) biotechnology company. Instead, the Department of Health and Human Services last April funded only one cell-based flu vaccine - $97 million for a vaccine that has not yet been tested in animals or humans.
"I don't know what the hell they are thinking about," said Dan Adams, president and chief executive officer of Protein Sciences. . ."
This is the voice of a man whose company missed out on a $97 million dollar contract, so that has to be taken into account. But it does appear that HHS and the FDA have been overly cautious about moving to cell-culture based vaccines.
But "caution" is a popular word in the vaccine field, in the financial, medical, and legal senses. and that brings up another provision in the President's proposal that I haven't seen anyone else comment on yet: liability protection for vaccine producers. As you might figure, I think that this is on principle a good idea, but the trial lawyers (and some others) will think differently. This will be an interesting fight, but it might take place largely out of sight. "Fight for your right to sue the people who are trying to protect you from bird flu" isn't a very catchy slogan.
My last comment on vaccines in this context is to point out that - cell culture or no cell culture - if we get to the point that we're relying on a vaccine to save people from a pandemic, then we could be in big trouble. There's an inevitable delay in vaccine development and production - months and months and months of delay, and that's when things are really zipping along. Viruses can mutate in the time it takes to fight their previous versions. If we're lucky, the vaccines that are being developed now will have enough protective effect against whatever flu strain might cause a pandemic. But they might well not, and we need to realize that.
+ TrackBacks (1) | Category: Infectious Diseases
November 1, 2005
I mentioned an interesting paper that's coming out in the Journal of Medicinal Chemistry on molecular modeling. It's a long one from a large group of people scattered across GlaxoSmithKline's worldwide research facilities, entitled "A Critical Assessment of Docking Programs and Scoring Functions." And that's what it is, all right.
For the non-med-chem readers, those are two of the key techniques in computational molecular modeling. Docking refers to taking a modeled version of your small molecule and trying to fit it into a similarly modeled version of the binding site of your protein target. The program ties to take into account the size and shape of the molecule and the binding site, of course, as well as more subtle interactions between the various functional groups. Scoring functions are what the programs use to try to rate how well the docking procedure went for a given compound, and to compare it to others in a given data set.
The GSK team did a very thorough job, evaluating ten different docking programs. They started with seven varying types of protein targets, mostly different classes of enzymes, all of which are known drug targets. An expert computational chemist took each one and polished up the model of the binding site. At the same time, lists of between one and two hundred potential binding compounds were put together for each target, including several series of related compounds. Another modeling chemist took these structures and got them ready for docking. They made sure that a crystal structure of each structural class was known for each case (to check the accuracy of the modeling later on), and also made sure that the binding affinity of the compounds ranged over at least four orders of magnitude (from pretty darn good, in other words, to pretty darn awful). The goal was to make the whole exercise as real-world as possible. Then each of those binding site models and their associated lists of potential ligands were turned over to separate chemists with experience in the various docking programs, and they told them to have at it. As the paper puts it:
"To optimize the performance of each docking program, computational chemists with expertise in a particular program were identified from the worldwide GSK computational chemistry community. Each program expert was given complete freedom and sufficient time to maximize the performance of the docking program. . .No time deadlines were imposed so that even low-throughput docking programs could be evaluated. Indeed, no constraints whatsoever were placed on the level of agonizing over details of how each docking program was applied."
It's important to remember that the results of this paper come from experienced users who had a great deal of knowledge about the targets, and all the time they needed to mess with them. The aformentioned agonizing was devoted to three typical kinds of question that such software is designed to answer: The first was: what is the conformation (the 3-D physical "pose") of a small molecule once it's in a binding site? This is why they picked all these things with known crystal structures, since those provide a check with real data. Results of this test were OK, in some cases fairly good. Some of the target proteins seemed to have binding sites that were more suited for the capabilities of the programs, which could take the majority of the compounds in their list and fit them pretty close (within two angstroms) to the known crystal structures.
And every target had at least one program that could take at least a third or so of the test compounds and dock them fairly well. But the problem was, no one program could do that for more than 35% of the binding modes. The best performances were scattered among the different software packages, and there seems to be absolutely no way to know in advance whether a given program is going to perform well on a new target. The other problem, and it's a big one, was that the scoring functions couldn't reliably identify when the program had hit on one of the good answers. There wasn't much correlation between what the program thought was a well-docked conformation and its resemblance to the known crystal structure.
The second question they looked at was: given a list of molecules (some active, some inactive), how well can the software pick out some active ones? This process is often known as "virtual screening". Again, the results were fairly good, but with some significant problems. For all but one of the targets, at least one of the programs could find at least half of the top 10% of the active compounds. (I know, that sounds like a lot of defensive hedging compared to what some people think these programs can do, but that's the real world for you). The programs also did pretty well at pulling a variety of structures out, and not just making their total by grabbing only the members of one particular class.
But that fairly-decent performance is for the programs as a group. As before, though, the best performances were scattered through all the software packages, with no real standout. Most of the programs, at one point or another, had to grind through a significant amount of a compound lists to do the job, too, which is something you really don't want in real-world use. Another disturbing result was that some of the scoring functions seemed to be picking the right compounds for the wrong reasons – that is, based on incorrect binding modes.
Now we're ready for the third question, a hard one which (in my experience) is one of the ones that medicinal chemists most would like molecular modeling software to answer: given a list of compounds, can the program rank-order them according to their expected affinity for the target? Unfortunately, the answer is "absolutely not." No scoring function in any of the software packages could even come close. The compounds that the programs ranked as winners were just as likely to stink, and the ones that they put into the discard heap were just as likely to be fine.
My way of looking at the first two tests is to say that if you have just one molecular modeling package, it is guaranteed to mislead you a fair amount of the time. And you have no way of knowing when it's doing that. If you have more than one program to work with, though, then they are guaranteed to disagree with each other a fair amount of the time, and you have no way of knowing which one of them is right – if either. I'll let the authors have last word on the third test, and on the software in general:
". . .in the area of rank-ordering or affinity prediction, reliance on a scoring function alone will not provide broadly reliable or useful information. . .This study demonstrates unequivocally that significant improvements are needed before compound scoring by docking algorithms will routinely have a consistent and major impact on lead optimization. . .it is not completely obvious by what means these improvements will arise. . ."
+ TrackBacks (0) | Category: In Silico