Nanopore Basecalling for Directed Evolution

on YouTube ↗28m

" summary

Derives a systematic approximation that turns an infinitesimal indel generator into a time-dependent pair HMM with nine transition probabilities reduced to four coupled differential equations, solved numerically (Runge‑Kutta). Uses moment‑matching / coarse‑graining by multiplying finite‑time transducers and collapsing states, improving relative‑entropy fit versus prior heuristics (including De Maio and TKF91 limits). Produces gap‑probability formulas that plug into automatic differentiation and can be fit to alignment data to estimate context‑dependent indel rates for proteins and genomes.

AI-written summary · about

" chapters (4)

1. 0:00Start
2. 0:25Why model indels phylogenetically?
3. 6:47Thorne, Kishino & Felsenstein (1991)
4. 9:51Continuous-time Markov processes Moment evolution approach

" related talks

• 27mA Statistical, Reference-Free Algorithm Subsumes Myriad Problems in Genome ScienceJulia Salzman · From Algorithms to Discovery in Genome-Scale Biology and Medicine 2022
• 53mHow machine learning is influencing protein engineeringJennifer Listgarten · Unknown Futures of Generalization 2024
• 01h 20mIntroduction to Regulatory Genomics and Epigenomics II: Motif FindingYaron Orenstein · Algorithmic Challenges in Genomics Boot Camp
• 38mInterpreting Deep Learning Models Of Functional Genomics Data To Decode Regulatory Sequence...Anshul Kundaje · Interpretable Machine Learning in Natural and Social Sciences (2022)
• 01h 20mFundamental Algorithms in Deep Sequencing II: Mapping/AlignmentLior Pachter · Algorithmic Challenges in Genomics Boot Camp
• 01h 13mDeep Learning in Structural Biology and Protein Design: How, Where, and WhyChloe Hsu · Deep Learning Theory Workshop and Summer School 2022
• 46mMachine Learning-based Design of Proteins and Small MoleculesJennifer Listgarten · Emerging Challenges in Deep Learning (Simons Institute workshop, 2019)
• 53mApplied Functional Programming in Computational GenomicsJamie Dixon · NDC Oslo 2022