search

CshlSiepelLab / phast

PHAST
Other
67 stars 24 forks source link

Problems with estimating rho #65

Open kuangzhuoran opened 10 months ago

kuangzhuoran commented 10 months ago

my command line was: "msa_split Chr1.maf --in-format MAF --windows 1000000,0 --out-root split --out-format SS --min-informative 1000 --between-blocks 5000" "phastCons --estimate-rho $ss --no-post-probs $ss nonconserved.mod"

Then I met: " WARNING: likelihood decreased during EM: it 162851 total_logl=-298744.3233, prev_total_logl=-298744.1436 WARNING: likelihood decreased during EM: it 162853 total_logl=-298744.4031, prev_total_logl=-298744.1678 WARNING: likelihood decreased during EM: it 162855 total_logl=-298744.3233, prev_total_logl=-298744.1436 WARNING: likelihood decreased during EM: it 162857 total_logl=-298744.4031, prev_total_logl=-298744.1678 WARNING: likelihood decreased during EM: it 162859 total_logl=-298744.3233, prev_total_logl=-298744.1436 WARNING: likelihood decreased during EM: it 162861 total_logl=-298744.4031, prev_total_logl=-298744.1678 WARNING: likelihood decreased during EM: it 162863 total_logl=-298744.3233, prev_total_logl=-298744.1436 WARNING: likelihood decreased during EM: it 162865 total_logl=-298744.4031, prev_total_logl=-298744.1678 WARNING: likelihood decreased during EM: it 162867 total_logl=-298744.3233, prev_total_logl=-298744.1436 WARNING: likelihood decreased during EM: it 162869 total_logl=-298744.4031, prev_total_logl=-298744.1678 "

It seems to have been calculated something over 100,000 times. I've had a quick look and in those 100,000 calculations, 'total_logl=' and 'prev_total_logl=' are just 2-3 numbers repeating back and forth, without any other changes

The problem occurred on multiple chromosomes, and this is one of the files that had problems: Problematic.ss.file.txt

navyashukla-uom commented 6 months ago

Hello, did you ever figure out what was happening? I'm having the same issue.