<scribe> ScribeNick: Michel

<scribe> Agenda: CPIC

CPIC is the clinical pharmacogenomics implementation consortium

bobf: member of CPIC

<BobF> http://pgrn.org/display/pgrnwebsite/PGRN+Home

bobf: PGRN - pharmagcogenomics research network - 13th year of funding, well organized and funded
... relates to PharmGKB and CPIC
... lead discovery and translation in pharmacogenomics
... PharmGKB is the associated kb for the PGRN; initially funded as part of PGRN. PGRN now just funds research sites and PharmGKB is spun out and has independent funding
... 14 research sites focused on specific drugs or diseases

<BobF> http://pgrn.org/display/pgrnwebsite/Research+Groups+and+Network+Resources

bobf: CGM - center for genomics medicine - sequencing collab with Riken
... DSR - deep sequencing resource
... PG-POP - EMR-linked biobanks; hypothesis testing
... PHONT - pharmacogenomics ontology - goal is to help bring standardized representation for data in PGRN
... bobF is a member of this
... shifting activities towards facilitating and demonstrating how PGx knowledge can be integrated in EMR in order to identify candidates for PGx studies

<BobF> https://www.pharmgkb.org/index.jsp

bobf: PharmGKB initially wanted to capture sequencing data, but competing with others (e.g. NCBI), since becoming independent have shifted focus away from collecting sequencing data, to capture knowledge for clinical action

<BobF> http://www.pharmgkb.org/projects.jsp

bobf: CPIC develops and publishes guidelines
... PharmGKB hosts the electronic version of the guidelines

<BobF> https://www.pharmgkb.org/resources/cpic_gene-drug_pairs.jsp

bobf: CPIC groups prepare guidelines
... based on genotyping that are approved for clinical settings

<BobF> http://www.ncbi.nlm.nih.gov/pmc/articles/PMC3098762/pdf/clpt2010279a.pdf

bobf: publication announcing the formation of CPIC
... guidelines have two schemes - genotypes into predicted phenotypes, phenotypes into drug dosing guidelines
... A - quality of evidence;

Level 1: the evidence includes consistent results from well‑designed, well-conducted studies.

Level 2: the evidence is sufficient to determine the effects, but the strength of the evidence is limited by the number, quality, or consistency of the individual studies, by the inability to generalize to routine practice, or by the indirect nature of the evidence.

Level 3: the evidence is insufficient to assess the effects on health outcomes because of the limited number of studies, insufficient power of the studies, important flaws in their design or in the way they were conducted, gaps in the chain of evidence, or lack of information

scribe: B - strength of recommendation

A: strong recommendation for the statement B: moderate recommendation for the statement C: optional recommendation for the statement

scribe: challenges and opportunities: have a nice format for publishing these, there is no consistent representation of this information that is machine readable
... opportunity for this group to make this machine readable and contribute back through PharmGKB
... groups are facing a challenge of biology - as more information is gathered (drug-disease-interaction) - landscape is getting more complex;

<ericP> warfarin dependent on 2 different genes

scribe: warfarin dosing depends on two genes; antidepressants have many genes involved, and the response is complicated

<ericP> tricyclics dependent on many more

scribe: so how to address this situation?
... current approach is not feasible in terms of scalability
... another challenge has to do with outcome data. Little data that genotype-based therapy is effective

<BobF> http://pgrn.org/display/pgrnwebsite/Network-wide+Projects

scribe: out of scope for CPIC, but in scope for translational pharmacogenetics project (TPP)
... funded to implement these guidelines internally, and develop and collect metrics to compare across sites
... the last 2 challenges/opportunities have to do with biological knowledge. Sequencing uncovering large numbers of variants
... all could impact previously published guidelines
... CPIC will have to revise guidelines; expanding the number of alleles that have a defined effect; (star alleles)
... another challenge - predicting the phenotype (functional effect) based on genotype; normally based on lab studies - errors in this and lack of precision

<ericP> predicting the functional diplotype based on phenotype

scribe: with multiple genes involved in a pathway - things get messy quickly - a kind of systems biology approach is required

<mscottm2> Regrets from Joanne - she wasn't able to schedule around this meeting and has Mark Musen visiting.

<ericP> michel: had a discussion at AMIAA about exponential guidelines to capture all of a patient's gene combos

<ericP> ... for making this machine-understanable:

<ericP> ... .. need to look at sophisticated approaches for retrieving phenotype data

<ericP> ... .. matthias has been looking at SPARQL Simple Rules

<ericP> ... .. also looking at OpenCDS

<ericP> ... where do we start the discussion about what's the best machine-readable format?

<ericP> BobF: PharmGKB isn't interested in "bit flipping"

<ericP> ... anything we do would need to be compatible with or translatable to existing commercial EHR systems

<ericP> ... if we can make CDS rules expressable in existing systems would be well-recieved

<ericP> ... revolution would be viewed as academic

<ericP> michel: at Mayo, what CDS is available to physicians?

<ericP> ... could rules be used there?

<ericP> BobF: discussing this with Mayo now

<ericP> ... looking at systems like OpenCDS

<ericP> ... looking at these options now

<ericP> michel: while we'll persue a SemWeb approach, we need to work with a partner who will apply them to an existing in-house system

ericP: does existing EMR systems capture the PGx information that a PGx CDS needs

bobf: there are no commercial EHRs that i'm aware of that directly supports genetic information
... degree of information varies

<ericP> genotype information captured under misc in existing systems

<ericP> ... you could put base pairs, alleles, genotype

bobf: rules can be tied to a drug
... so that an event pops up to the physician
... prespective / postpective -

ericp: not only not just capturing the information, but also not doing the assessment of effectiveness

rich: how population specific are these guidelines? in the elderly you might suggest a lower dose, but these person has 5 other issues;

bobf: CPIC is specific on genotype-guided therapy

<mscottm2> /me asks richard - what is an SSRI?

<boycer> selective serotonin reuptake inhibitor - popular antidepressant class

<mscottm2> tx

<ericP> SSRIs

<mscottm2> you guys are fast

bobf: going beyond genotype-based guidelines really needs CDS

<Yesterday> drugs from different locales (countries) will have different potency so must be taken into consideration.

<Yesterday> Bob knows me.

<Yesterday> I am from Duke

<Yesterday> I wasn't sure what my interest will be on this call so didn't want to properly register.

<Yesterday> This is my first call.

<Yesterday> Thanks for the welcome!

<egombocz> thanks, very informative

<mscottm2> Scott: Isn't it an uphill battle with pharmacogenomic info not on the radar for the clinic?

<mscottm2> http://lists.w3.org/Archives/Public/public-semweb-lifesci/2012May/0017.html

<mscottm2> BobF: There are dozens of clinics that are far ahead with this although most solutions are still ad hoc