Haplotype probability bug fix and changes to noise usage in expression estimation

jonassibbesen / rpvg

Method for inferring path posterior probabilities and abundances from pangenome graph read alignments

MIT License

47 stars 6 forks source link

Closed jonassibbesen closed 3 years ago

jonassibbesen commented 3 years ago

Improvements:

Removed noise transcript from EM algorithm. Instead a noise-adjusted read count is calculated. This gets around the noise having an expression value.

Changes to existing options and outputs:

Removed ClusterRelativeExpression from output. Can easily be regenerated by calculating the relative read count in each cluster.

Bug fixes: