1/ Genome-wide (and other) positive selection screens in genomes may be severely biased by alignment and other errors. we (@averyselberg, @NathanClark111, @ChikinaLab, @nekrut, @stvnwvr, @aglucaci, @timsackton) propose a simple fix.

The 15% indirect cost cap by the NIH is denotating a nuke under the current business model of many US schools, especially medical schools. Major changes incoming. Maybe relaized savings will translate to larger grant budgets/better success rates, but I not holding my breath.

Oops. On the day the new 2024 highly cited researcher list is released, clarivate website crashes. How 1990s.

7/ Google LLM "discussion" for the interested.

@aglucaci @NIH_LRP @theNCI @LabMelnick @mason_lab Congrats! Well deserved.

@NathanClark111 Better that than jet fuel. Wild!

RT @biorxiv_bioinfo: AUTO-TUNE: SELECTING THE DISTANCE THRESHOLD FOR INFERRING HIV TRANSMISSION CLUSTERS #biorxiv_…

@TrumanLab I think the original intention was that a single R01 could sustain a lab. A small lab, but a functional lab. That has been a fiction for >20 years now. My lab is small and I need at least 2 R01 equivalents. And that's with hard money for me. Soft $ people have it much worse.

@ClausWilke Hey, man, you had a great excuse for (any)/low turnout. The quality of your seminar was not identifiable.

@davorasmussen Nice work! Surprised at how decently GARD is doing, especially for the 3-seq dataset where the only information is a different branch length. Thanks for making the data available as well. I may play around with it some more.

@TeePi11 @DevoEvoMed @hyphy_software But we do have quite a bit of data on saturation in this paper (in the context of molecular dating). Branch-site type models actually help you rescue better estimates of branch lengths as well.

@TeePi11 @DevoEvoMed @hyphy_software In we show that this (as most commonly used models in PAML do) can lead to both positively misleading results and loss of power. dS saturation and CG bias are secondary concerns. This is a more fundamental issue, I think.

@nomad421 Always worth looking at Agner Fog's stuff ( Native width (64 bit) shifts are 1 op. With SIMD, you can do more, e.g. NEON ops like vshlq_n_s32 which does 4 at a time

@theosanderson Not sure these letters are more than a checkbox. Plus now they will mostly be written by ChatGPT or another LLM anyway.