#5 · Genomics

Sequencing and inference methods resolve nucleic acid sequence, structure, and usage

The dominant thread is quantitative inference from nucleic acid sequence-level measurements, in which an explicit model is fit to data and its parameters or structural states are estimated. A Bayesian framework recovers motif rate equation parameters from ligation counts in RNA reactor simulations and reports uncertainty estimates. An entropy-based index with linear models relates codon usage bias across 1,406 species to amino acid properties and the second codon base. sm-PORE-cupine couples chemical probing with direct RNA nanopore sequencing and signal alignment to resolve RNA structure ensembles on individual molecules. What is new is the resolution of ensemble heterogeneity and parameter uncertainty directly from count-level and single-molecule sequence data.