The genetics department at Washington University hosts a seminar every Thursday. The tradition of inviting outstanding scientists to speak goes back a long time(I have no idea when this started). As graduate students we get a chance to eat lunch and interact with the speakers from other institutions. I wanted to write briefly about some of the seminars I’ve attended in Fall 2019.

Rick Lifton gave the keynote for the Precision Medicine pathway retreat. Dr. Lifton had a bunch of interesting human genetics cases that he spoke about. Some of the cases Dr. Lifton described had parallels to X-men, except for the fact that his patients are real humans. One of them had an extremely low fracture rate due to his outlier bone density(many z-scores out). The story is described here. Another case that he talked about was someone who had to constantly drink pickle juice to maintain appropriate salt levels. This research is described here . The genetic causes of his condition helped shed light on genes involved in and mechanisms of hypertension. This strategy of identifying outlier individuals for some phenotype, checking if the trait is observed in several members of the individual’s family and then narrowing down on the genetics behind the trait is now a strategy adopted by several human genetics groups. The UK BioBank and similar biobanks elsewher(Japan, Finland, Estonia and Vanderbilt come mind first) seems well positioned to identify such cases. Other resources are actively being setup, I learnt about a similar resource being developed at Stanford currently with data from wearables being actively collected. In discussions with my advisor it’s a bit strange that more of the medical institutions in the US haven’t raced to do this with the EHR data available. This is happening behind the scenes right now and the countries with centralized health systems seem to be ahead of the rest at the moment.

Hunter Frasier from Stanford gave what I found to be one of the most intriguing seminars of this semester. Dr. Frasier’s group and collaborators used an approach called CRISPEY to precisely induce single nucleotide changes in yeast strains. This is described in this paper This approach could be helpful to identify the causal nucleotide change in a QTL. QTLs are usually loci that encompass a lot of genetic variants and identifying the causal variants is still an open problem. Dr. Frasier’s group tested variants present in extant population. The results were super interesting, most of the changes with larger effects were non-coding SNPs instead of coding SNPs. One way this result could make sense is if the large effect coding SNPs had already been weeded out by natural selection in extant populations, this makes sense. However the strange part was the causal QTNs showed little overlap with the loci detected by traditional crossing methods. This is strange and potentially very interesting.

There were several other seminars that I was able to attend and found interesting. Eimear Kenny gave a super interesting talk on the population genetics of NYC. The stories from this cohort were very interesting including a SNP predicted to have a huge effect on height(I think its 6 - 10 inches) They also developed newer faster methods to detect IBD in large cohorts. Some of the results from the talk are in this paper. This is one of the most diverse cohorts in the United States at the moment and I imagine working with these samples must be super exciting. The lunch with Dr. Kenny was super interesting as the students touched on different issues including genetic privacy. At a later seminar, Stephen Montgomery spoke about the state of co-localisation approaches for GWAS variants. Most variants tested in GTEx look like eQTLs in at-least one tissue. So testing if a GWAS variant is also an eQTL is not a simple lookup and can lead to a false-positives. Instead sophisticated statistical methods which test if the structure of the signal in the GWAS haplotype matches the structure seen in the eQTL signal are now the norm. However the problem linking a GWAS hit to a gene still still isn’t solved. Dr. Montgomery also spoke about identifying rare variants in gene expression outliers and how this information could potentially be used to improve PRS predictions. This reminded me again of the human genetics approach of identifying “all the outliers”. Jennifer Phillips-Cremins gave a super interesting seminar on how 3-D genome structure is altered at the breakpoint of repeat expansion disorder patients. This change in the TAD structure could potentially alter transcriptional regulation or could be the result of such an alteration. Dr. Phillips’ group has some amazing technology for creating loops between loci, this could be used to examine the downstream effects of loop formation. Sabrina Spencer gave a fascinating about non-genetic hetrogeneity and tracing down why all cells don’t cycle through the cell-cycle in an identical manner. Dr. Spencer’s group has developed a live-cell sensor to monitor CDK2 levels and used this to identify the cells which don’t progress through the cell cycle. Some of these are found to be cells with likely DNA damage. This area of research is pretty fascinating and could potentially uncover a lot of important cell biology.

I found the Fall 2019 edition of the Genetics seminar series to be interesting. The department seminars could potentially be an important part of graduate student training. I hope the tradition continues in the following years.