Brian L Trippe
@brianltrippe
Followers
2K
Following
436
Media
12
Statuses
108
Assistant Professor at @Stanford Statistics and @StanfordData | Prev postdoc @UWproteindesign and @Columbia. PhD from @MIT.
Stanford, CA
Joined December 2016
š„ Benchmark Alert! MotifBench sets a new standard for evaluating protein design methods for motif scaffolding. Why does this matter? Reproducibility & consistent evaluation have been lackingāuntil now. Paper: | Repo: A thread ā¬ļø
5
41
135
RT @arthurdeng0205: Predicting how mutations affect protein binding is key for drug designābut deep learning tools lag behind physics. Staā¦.
0
2
0
RT @Zhuoqi_Zheng: Interested in functional protein design or motif-scaffolding? We propose a standardized protein design benchmark, MotifBeā¦.
0
8
0
RT @DidiKieran: Have a look at our shiny new benchmark for motif-scaffolding in computational protein design! New (and harder) tasks, incluā¦.
0
4
0
RT @KevinKaichuang: We made a new, reproducible, fair, (and much harder) motif scaffolding benchmark!. With Zhuoqi Zheng, Bo Zhang, @DidiKiā¦.
0
11
0
And thanks as well to ChatGPT for turning my first banal attempt at a tweet-thread into a emoji-packed click-bait for real work! š.
0
0
8
This has been a big team effort ā with contributions across 5 timezones š. This is thanks to my fantastic coauthors on our whitepaper:
1
0
2
This is a V0 pilotāwe need your input for V1!.š§© Know of an important motif? Add it to the benchmark!.šļø Help improve the pipeline & metrics (sequence-based? Side-chain-level?).Letās shape the future of motif scaffolding together!.
1
0
3
Surprise: A modern baseline (RFdiffusion) fails on motifs scaffolded into de novo enzymes 15 years ago š¤Æ.This suggests modern deep learning methods arenāt always better than past methods, and there is much room for progress!
2
0
6
MotifBench fixes this by providing:.š§Ŗ A standardized evaluation pipeline.š 30 challenging motifs as test cases.š Easy-to-use eval scripts and a leaderboard for method comparison.Now, results can be easily and consistently measured.
1
0
2
Recent progress in motif scaffolding has been exciting! Butā¦.ā Current evaluation are inconsistent, and results incomparable.ā Widely used test cases are too easy.ā Reproducibility is difficult.Thatās where MotifBench comes in.
1
0
1
Motif scaffolding is a core challenge in protein design:.ā
Input: a motif (small functional substructure).šÆ Goal: find scaffolds (full proteins) that preserves the motifās geometry. But what's the state of methods for this problem?
1
0
3
RT @MichaelOberst: I'm recruiting PhD students for Fall 2025! CS PhD Deadline: Dec. 15th. I work on safe/reliable ML and causal inference,ā¦.
0
125
0
I am pleased to share that I have accepted a position @Stanford to start as an assistant professor in the department of statistics, with an affiliation in @StanfordData, this fall!.
82
27
891
Come check out our poster today at 4pm! We find that with a sequential Monte Carlo technique known as twisting we can use conditional sampling heuristics, like reconstruction guidance, to get exact inferences with additional compute.
Twisted Diffusion Sampling - Practical and Asymptotically Exact Conditional Sampling for Diffusion Models. Theoretically grounded conditional sampling for diffusion models with applications to inpainting, Bayesian inverse problems, protein design, .
0
3
25
RT @renberlinghieri: Where better than Hawaii to talk about oceans?š Excited to present our work on ocean current modeling using Gaussian pā¦.
0
7
0
RT @json_yim: I'm at ICML next week presenting FrameDiff in main conference and our regularized protein optimization method in the workshopā¦.
0
7
0
RT @HelenEisenach: Just a reminder that RFdiffusion is free and publicly available, and thanks to @sokrypton, accessible through Google Colā¦.
0
26
0