DOPseq -> SnakeMake - Githubissues

ilyakichigin / DOPseq_analyzer

0 stars 1 forks source link

DOPseq -> SnakeMake #15

Open ilyakichigin opened 6 years ago

amakunin commented 6 years ago

Feature requests:

[x] single conda env
[x] per-sample reference and filtering setting
[x] regions should cover entire genome - we classify it, not select some parts
[x] add position log-ratio and associated one-sided p-value (log-ratio > 0)
[ ] MultiQC
[ ] positions statistics
[ ] validate single reference genome per sample
[ ] update readme with log-ratio and p-value concepts
[ ] deduplicate trimming/alignment of same reads/same reference

amakunin commented 6 years ago

Python notebooks showcasing downstream analysis

Chromosome visualization ipynb:

[ ] replace data with something more meaningful
[ ] subsetting and outputting chromosomes demo

Regions visualization ipynb - interactive alternative to DNAcopy plots. Ideas - plot n chromosomes with significant regions, compare samples. Also - subsetting and outputting region data.

amakunin commented 6 years ago

Region calling potential bug: log_ratio and p_value are calculated for region length. Thus, a few high density positions in the middle of the scaffold are "expanded" to the entire scaffold size, which reduces signal strength.

Also, pd_mean is taken directly from DNAcopy output, so it is not 100% accurate estimate.