Fundamental Algorithms in Deep Sequencing II: Mapping/Alignment
summary
Distinguishes classical sequence alignment (inferring evolutionary homology forests) from read mapping (short fragments with sequencing error), and argues most pipelines should stop producing full base-level alignments. Reviews Burrows–Wheeler/FMI-based tools (Bowtie, BWA) and BLAST, then builds a fast pseudo-alignment using a De Bruijn graph of k‑mers (k=31), implemented as Callisto/Kallisto, that returns equivalence‑class counts and lets EM assign ambiguous reads. Claims terabase runs fit on a laptop (roughly 4–5× input size) and highlights trisomy‑21 plasma sequencing and RNA‑Seq as primary use cases.
AI-written summary · about
chapters (6)
- 0:00Start
- 1:16A topic that doesn't exist
- 14:35Example of a homology forest
- 16:48What is an alignment?
- 31:27A different type of sequence alignment problem: read alignment
- 51:04Measuring by sequencing (sequence census)
related talks
53mApplied Functional Programming in Computational Genomics
27mA Statistical, Reference-Free Algorithm Subsumes Myriad Problems in Genome Science
38mPython in Bioinformatics
11mIntroduction to Smith-Waterman – An accurate DNA sequencing algorithm
31mParallel Biological Sequence Comparison in Large GPU Clusters
31mHacking the Human Genome Using Clojure and Similarity Search
28mSome Current Computational Challenges in Bioinformatics
01h 13mDeep Learning in Structural Biology and Protein Design: How, Where, and Why