Sign in

Richard Neher

@neher.io
703 followers 273 following 37 posts
PostsRepliesMedia
Richard Neher @neher.io · 12/04/2026
Thank you, Hungary!
041
Richard Neher @neher.io · 23/08/2025
Within the syntenic core genome, linkage between SNPs decays rapidly with distance. LD approaches background levels after about 1000 bases. The background level itself is often set by population structure with little linkage within subgroups but genome wide coupling across subgroups. [4/6]
Linkage decay in phage core genomes.Population structure increases background levels of SNP linkage.
220
Richard Neher @neher.io · 23/08/2025
Phages are known to recombine with each other and have flexible and fluid genomes. We analyzed diversity within the clusters of the Acinetobacteriophage Database by organizing their genomes into pangenome graphs based on homologous phams (protein families). [2/6]
construction of pangenomes graphs of phage genomes
162
Richard Neher @neher.io · 16/03/2025
Numbers came down quickly in early April, but the excess deaths accumulated through March corresponded roughly to the number of people that normally die in half a year.
010
Richard Neher @neher.io · 16/03/2025
5 years ago, I subscribed to Eco di Bergamo and spent a few hours counting obituaries in Feb/March. Most cases in northern Europe at the time were travel returnees that didn't get very sick, but the 7x increase in mortality in Bergamo left no doubt what happens if the virus hits the vulnerable.
Number obituaries in Bergamo in March 2020.
130
Richard Neher @neher.io · 11/01/2025
We use the improved estimates of mutation rates to identify regions where synonymous mutations are under purifying selection. Most clear signals correspond to known structures like the ORF1a/b frameshift and TRSs. But two clear signals lack an explanation (in E and between M/ORF6). [7/N]
Selection on synonymous and non-coding sites in SC2
101
Richard Neher @neher.io · 11/01/2025
A simple linear model with genomic region, 5' and 3' neighborhood, and 2nd pairing explains between 15 and 60% of the fold-variation of the rates. [6/N]
Variance in log-mutation rate explained by a linear model.
111
Richard Neher @neher.io · 11/01/2025
The underlying mechanism is a mystery to us (we speculate a bit in the preprint). Rates also depend on neighboring bases, sometimes by more than 10-fold. These neighbor-dependence is very strand symmetric for some mutations (e.g. T>G and A>C), but not others. [3/N]
Context dependence of SC2 mutation rates.
111
Richard Neher @neher.io · 11/01/2025
Some rates, like C>T(U), are much higher than others. But even for the same mutation type, rates vary dramatically from site to site, sometimes in striking patterns: A>T mutations are 2-3 fold more common after the start of Spike. C>T drops almost two-fold at the ORF1a/b boundary. [2/N]
SC2 mutation rate variation along the genome.
211