findMotifs vs findMotifsGenome, inequality of results


Hi all,

so I'm testing Homer for motif analysis. There are two tools

  1. which search motifs in "promoters" and also in provided fasta files (I don't want to use the tss mode, because I want to analyze also intergenic features with this tool).
  2. which search for motifs using preconfigured genome (fasta plus gtf) and sequences in bed files.

Both programs share majority of options, the main difference, as I see, is lack of -size option in findMotifs.

Maybe it causes differences when analyzing the same sequences by both methods, here are my scripts: peaks_factor/noctg_prom_ssc.fa zea findmotigs_ssc -fastaBg peaks_factor/noctg_prom_ss.fa -nomask -mset plants -mcheck ~/bin/homer/data/knownTFs/plants/all.motifs -mknown ~/bin/homer/data/knownTFs/plants/known.motifs -p 4 -fdr 100 peaks_factor/noctg_prom_ssc.bed zm_v4old anal_homer_diff -bg peaks_factor/noctg_prom_ss.bed -h -size given -mcheck ~/bin/homer/data/knownTFs/plants/all.motifs -mknown ~/bin/homer/data/knownTFs/plants/known.motifs -nomask -p 4 -fdr 100

The output of reports 732 "experimental" sequences and 293 control sequences.
Whereas for the corresponding values are 732 and 123.
I really don't know why..., as I have 732 "experimental" sequences and 338 control ones.

Both scripts give different results, does anyone have any experience with that scripts and can advice, which one is better?
Here is the link to documentation.




Source link