Stanford offers cool class to med students: analyze your (or someone’s) genome!

I am very pleased that my colleagues at Stanford have decided to offer a class on personal genomics to the medical students.  The class is elective, and getting your 1 million SNPs tested is optional.  If you want your own SNPs, you have to pay $99 (it’s sounds like a copay, why not also give them the experience of our whacky medical finance system while we are at it?  They are going to learn a ton in this course!).  If you don’t do your own genotypes, they have a standard reference genome for you to do the homeworks.  The class is lead by my colleague Stuart Kim and an enterprising MD/PhD student, Keyan Salari, who really fueled the effort to offer the course.  This course was not easy to get going.  There was significant and legitimate concerns among the faculty about the prudence of doing this.  There was an entertaining and informative email debate during which we fleshed out the issues.  We also watched the response to the proposal to genotype incoming Berkeley undergraduates (I’m also a fan of that).   The Stanford Dean’s office has released an announcement discussing the course and the decision process and safeguards that have been put in place.  Outstanding.

I am honored to be participating in the class as a lecturer on–you guessed it–Pharmacogenomics.  The students have all heard my standard talk as part of their basic genetics curriculum, so we will do some advanced stuff in this class.  I think that we will have them assess the genetics of response to some drugs based on very solid pharmacogenomics evidence:  statins, clopidogrel and warfarin.  I will have them all compute the ideal dose of warfarin from the genome they are working with based on the dosing equation we published last year (or maybe the modified one we published this year) that uses genetics.   Then, for fun and to get them really thinking, we will assess the genetics of response to other drugs where the evidence is published but not as firm.  This is going to be the more common situation for using genetics in pharmacology, and they need to start understanding how to take imperfect evidence and fold it into their medical decision making about prescribing.  I haven’t decided which drugs to cover there, but our recent paper analyzing Steve Quake’s genome offered some interesting inferences on more than 100 drugs, so I will pick some that may be relevant to young physicians.

Anyway, this is great, and I am proud that we are trying to push the agenda of bringing personal genomics to medical training.   Someday, I hope this is a mandatory part of pharmacology or genetics or both, but I’ll take an elective at this point.


I can’t believe I’m saying this: Go Cal.

UC Berkeley recently announced that they will offer incoming freshman free genotyping at three genetic loci for folate, alcohol and lactose.  This is part of a tradition at Cal to engage entering students with some shared intellectual activity (the “On the same page” program).  I believe it is absolutely critical that we immediately start educating students about genetics, and the use of genetic tests in making decisions in life.  I think that direct engagement with personal data is one very effective way to make all the issues crystal clear.  This should be optional and there should be adequate safeguards, but this should not stop us from getting started in educating the general population about the promise and pitfalls of genetics.   Thus,  I think that this is great and that Cal (a traditional rival of Stanford, thus the title of this post) deserves credit for bringing genetics to the consciousness of our next generation of leaders.  Some have voiced concerns, and these can be handled with reasonable precautions.  I am particularly amused by the concern that knowledge of genetics might induce poor behavior with respect to alcohol consumption.  While this is certainly a theoretical possibility, it is my impression that students are already making poor decisions about alcohol consumption at such a rate that knowledge of genetics is unlikely to affect this trend appreciably.  In fact, anything that gets the pros/cons of alcohol consumption into the discussion while students are sober is probably a good thing.   We have featured alcohol genetics on a site to help high school biology students can see genetics in action.    So, at least in the NARROW ARENA OF PROVIDING GENETIC INFORMATION SO STUDENTS CAN LEARN…Go Cal!

Year in Review: 2009 Translational Bioinformatics

I presented my review of the year in translational bioinformatics at the AMIA Summit on Translational Bioinformatics.   It is highly biased and subject to all the problems of one person trying to do something like this, but I have made a PDF of the slides available here.  These are papers published since January 2009 to present.   I apologize for important papers that I have missed, and thanks to those who provided advice.   As I have mentioned previously, you can also see the 2008 version and the 2009 version.

Way to go CVS Caremark!

I just learned that CVS Caremark has decided to put some pharmacogenomics testing in place for a variety of clinical areas, including cancer, cardiovascular disease and HIV.   This is an alliance with Generation Health.   I am a CVS Caremark customer (thank you, Stanford), and I am thrilled to learn that they are doing this.   It raises the possibility that PGx penetration will happen not by FDA decree or government reimbursement agencies, but by pharmacy benefits companies that see real value to them and their customers.  It will be very interesting to see how this goes.   CVS Caremark is a big company that serves 60 million people.   If their competitors follow suit, we may see the benefits of pharmacogenomics reaching patients much sooner than might occur if we wait for other routes to materialize.

FDA looking for leads in pharmacogenetics.

I was excited to see the announcement of the relabeling of clopidogrel (a.k.a. Plavix) by the FDA.   The FDA exerts much of its influence on drug companies and medical practitioners by approving the “label” that describes the uses, doses, cautions, and much other information about using a medication.  It is the folded up piece of paper that is always inserted into your medication packaging (thus it is also known as the “package insert”), that then unfolds into a huge poster-sized document with tiny font and lots of information.  It is a legal document which specifies what the drug company can claim about the drug, and how physicians are supposed to use it (give or take some “off label” uses, but that is another topic).   I’m sure it is the star of many lawsuits.

In any case, there has been accumulating evidence that certain genetics variants of CYP2C19 that reduce its activity are associated with differential success of clopidogrel at preventing myocardial infarctions (heart attacks).  Also, drugs that inhibit CYP2C19 also reduce the effectiveness of clopidogrel.  Thus, the FDA modified the label for clopidogrel and warned about concomitant use of medications that interact with CYP2C19 and also mentioned that the genetic variants present in a patient may warrant careful assessment of whether to use the drug, and how much to use.  This is another success for pharmacogenetics!

The pharmacogenetics database we are developing,  PharmGKB, is focused on gathering evidence for the role of genetics in drug response.  PharmGKB staff have  a good working relationship with some FDA scientists, and we are all very interested in finding out about other drugs for which there is mounting emerging evidence that genetics may play an important role.  In some cases, the FDA may even launch a process to determine if the label should be modified in light of new evidence.  If you come across compelling scientific evidence for genetic determinants of drug response, let us know at  We will gather together responses and assess them on a regular basis–we will even alert our friends at the FDA to those that are reaching a critical point!

You can see a list of drugs with specific FDA language about genetics on PharmGKB.