Ling Liu’s beautiful PhD work on modeling transcriptional elongation and its genomic and epigenomic determinants is now published. A technical tour de force with a number of interesting biological observations

The abstract deadline for ProbGen '25 has been extended to January 24. Turnout has been good so far but we're hoping to increase it a bit more, so there's still time to submit if you have an idea for a talk or poster!.

Generate videos in just a few seconds. Try Grok Imagine, free for a limited time.

Please note that the abstract deadline for ProbGen '25 is this Fri, Jan 10. Please join us. We have a great line-up of keynote speakers and session chairs.

RT @Nowak_Lab: We are looking for a passionate #Postdoc to join our team in NYC @WeillCornell! Interest in #ProstateCancer, #Evolution, #Me….

ProbGen attendees: watch for Mo's talk next week!.

Latest preprint from our group, on a clever approach devised by Ziyi Mo to address a critical limitation of simulation-based deep-learning approaches in popgen.

Reminder that the abstract deadline for ProbGen '23 is one week from today!

RT @david_a_knowles: @asiepel and I are hosting our first post(?)-pandemic in-person pop/statgen seminar tomorrow 4pm at @nygenome with @pa….

RT @lizzyzhao: We are excited to announce that New York area population genomics meeting is back. The next meeting is on January 27, 2023,….

It was a good reminder that wet-lab experiments aren't the only ones that matter. Experimentation by simulation can be extremely informative and valuable! 14/14.

@YixinZhao1989 One of the most fun and satisfying parts of the project was iterating between simulation and model development. Using Yixin's simulator, we kept discovering new aspects of the process to incorporate into the model. 13/14.

@YixinZhao1989 was the lead on this project and did a massive amount of work to bring it over the finish line. Also joint work with Ling Liu, who has a related paper in prep (coming soon!) 12/14.

While the steady-state case is surprisingly informative, work is in progress to generalize the model to the nonequilibrium setting with time course data. 11/14.

By applying the model to real data, we find that steric hindrance substantially limits initiation rates, esp. during response to cellular stress. We also show that the location and variance of pause sites correlates with particular sequence motifs and stress responses. 10/14.

Along the way, we derive a precise mathematical relationship between the widely used "pausing index" and the actual rate of pause-escape under our model. 9/14.

Two aspects of the process turn out to be critical to include in the model to obtain a good fit to simulated and real data: variability across cells in pause sites, and steric hindrance of initiation by paused polymerases. 8/14.

We focus in particular on the interplay between these two processes, initiation and pause-escape. We show that they exist in a kind of dance, with each process limiting the other under certain circumstances. 7/14.

We show that, with careful statistical modeling, it is possible to learn a surprising amount from these steady-state data, including absolute rates of initiation and promoter-proximal pause-escape (after calibration). 6/14.

However, this information is challenging to extract: it's a bit like trying to work out the dynamics of cars on a highway by looking at patterns of headlight brightness from a distant hill. 5/14.

At the same time, even standard NRS data collected in a single time-point at steady-state contain rich information about transcriptional dynamics. And for all genes often without chemical treatment 4/14.