#40 · Genomics
Scalable inference of phylogenetic networks and repetitive sequence evolution
The dominant thread across these items is scalable inference from phylogenetic and sequence data that conventional methods handle poorly. CAMUS estimates level-1 phylogenetic networks from rooted constraint trees and quartets, reported faster than PhyloNet-MPL and scaling to 201 species. Phlag applies an HMM over quartet statistics from gene trees along the genome to detect regions inconsistent with the species tree. Three repeat-robust k-mer estimators quantify mutation rates between related sequences and are evaluated on highly repetitive alpha satellite sequences. The seed-chain-extend analysis extends average-case guarantees to indel channels, proving expected recoverability of at least 1-O(1/m) and bounded runtime. GBSC provides linear-time clustering of protein short tandem repeats. What is new is explicit handling of repetitive and graph-structured data at scale, although the items differ in data type and evaluation, and reported performance figures come from individual papers rather than a common benchmark.
Published papers
The gift of novelty: repeat-robust k-mer-based estimators of mutation rates.
Introduces three repeat-robust k-mer-based estimators of mutation rates between related sequences, evaluated on highly repetitive alpha satellite sequences;…
Scalable computation of ultrabubbles in pangenomes by orienting bidirected graphs.
Linear-time orientation algorithm for bidirected pangenome graphs enables linear-time ultrabubble enumeration, with up to 25x speedup over vg and over 200x…
CAMUS: scalable phylogenetic network estimation.
CAMUS is a scalable method estimating level-1 phylogenetic networks from rooted constraint trees and quartets; faster than PhyloNet-MPL, scales to 201 species…
VirBinn improves viral genome binning from metagenomic Hi-C through graph diffusion.
VirBinn is a graph-diffusion framework that enhances sparse virus-virus Hi-C contact graphs via random-walk-with-restart and host-guided diffusion, improving…
Phlag: scalable detection of genomics regions with unexplained phylogenetic heterogeneity.
Phlag is a scalable HMM-based method that uses quartet statistics from gene trees along the genome to detect regions inconsistent with a species tree beyond…
GBSC: graph-based sequence clustering method for similar short tandem repeats in protein sequences.
GBSC is a linear-time graph-based algorithm that clusters protein sequence fragments by short tandem repeat patterns, tolerating insertions, mutations and…
Topic history
| Date | Rank |
|---|---|
| #40 | |
| #33 | |
| #33 | |
| #38 |