Here's my talk at the 1st 🏳️‍🌈LGBTQI+ @ISCB Symposium, which was held on Monday online. It's about how coming out intersected with my education and scientific career. It's quite personal so if that's not your thing, skip it. Other talks on the ISCB channel.

The funniest thing on twitter today. This guy learned one thing in high school (redox reactions) and thinks everything in life is just that. I did something like that once. I was in 10th grade. (The alternative is he knows this is all BS and it is just a grift).

https://t.co/HrMF44wsqf

Oop, this is why they are doing this today:

In light of James Watson's death, I post this article which clarifies Rosalind Franklin's contributions to the discovery of the structure of DNA. It is richer than the conventional narrative. https://t.co/BW5abtUUba

🧵 NIH Director Jay Bhattacharya is speaking at Together's 4th Anniversary event. The group's co-founders include an anti-Semite and an anti-vax homeopath, and the group is full of chemtrail nuts and other conspiracy theorists. Here are some of the other speakers at the event...

I don't do much science here because X is a cesspit of hate. It's gotten worse in the last few weeks, because it seems the algorithm changed and just shows me tweets from Trump administration officials first. I don't want to hear from DUI hire @SecWar Pete Hegseth (now blocked).

all ideal place pictures reveal that everybody hates cars and probably don't even realise it

@SenMullin @WhiteHouse With his great sense of humor, 🍊🫡👔Donnie will definitely appreciate this snippet of the @Whitehouse Hall-O-Fame upcoming Halloween bash - where legends come to life - circa October 2025

So instead of using ipSAE_max or ipSAE_min, if the target is chain A and the design is chain B, use the "asym" line where Chn1 is "A" and Chn2 is "B". Chn1 is aligned, Chn2 is scored. Github: https://t.co/RToXL5IBXy I have some pull requests to address, which I'll do soon.

For designs, target is usually bigger, so aligning on target & scoring design means getting smaller d0, and hence minimum ipSAE. It seems logical to me to align on target and score the design. Evidence for this would be when target is smaller than design and use ipSAE_max... 2/3

Thanks for posting. I think the reason "ipSAE_min" works is bc ipSAE_min occurs when aligning on protein with more residues w/ PAE<PAEcutoff with respect to other protein & scoring on smaller. d0 comes from # scored residues (# w/PAE<PAEcutoff). smaller d0 -> smaller ipSAE... 1/3

WOW! ICE Nazis in Chicago tried to kidnap this food delivery worker but the guy was too fast for those ICE agents! 👏🏽🔥

@SecWar If you are willing to defend the slaughter of innocent civilians in the past, you are perfectly capable of ordering that in the present. You are a danger to this country and don't deserve your position. You're just a drunk sociopath.

My practical takeaways: 1. Prioritize by AF3 ipSAE_min; consider thresholds like ipSAE_min > 0.61, or use interaction filters such as ipSAE_min × (ΔG/ΔSASA) < −1.5, or LIS × shape‑complementarity > 0.42. 2. Pre‑filter designs with shape‑complementarity > 0.62 and optionally

ACIP Is Now Obsolete "the meeting devolved chaotically: a decision passed on Thurs by members who misunderstood the vote had to be reversed on Fri meanwhile, one panelist on a "hot mic" called another panelist an "idiot". 😳 You can't make this up. https://t.co/iPOcIV7DCH

Pay attention to James Talarico. He’s 36. He’s social media savvy. He communicates with all audiences calmly. He talks about issues not many Democrats talk about. He’s a star & he’s the future of our party—& he’s now running for Senate in TX. Watch this. https://t.co/EQrOMNpQdx

Predicting Experimental Success in De Novo Binder Design: A Meta-Analysis of 3,766 Experimentally Characterised Binders https://t.co/7CpAvJjsDl #biorxiv_bioinfo

Honored to receive the 2026 ASBMB Ruth Kirschstein Award for Maximizing Access in Science. @RTAatPennMed @Penn_CBIO.

Warren DeLano was a true hero of mine -- for producing wonderful software for looking at protein structure and distributing it for free (or nominal fees to support his work). Here's one of my favorite images i've made with PyMOL (AlphaFold models of active human kinases).

Pleased to receive the 2026 DeLano Award for Computational Biosciences. @ASBMB noted my LGBTQ+ Award from Temple Univ, enabling me to let young queer scientists that they're not alone & that there are successful queer scientists who've been around & out a long time (39 years!).