I have used findMotifsGenome.pl from HOMER to find motif, its output have many files including knownResults.html and homerResults.html. which output file should I consider for my downstream studies and what is difference between these two files?
homerResults is the de novo search, so the tool builds motifs from the data. What the output contains is the inferred motifs as logos, and then a closest match in the database you check against.
knownResults is an analysis in which it checks for the presence of known motifs in your sequences.
In well-annotated organisms like mouse and human the outputs are typically close to identical because many transcription factors and their motifs have already been identified. If an organism was poorly annotated then there might be a lot of de novo stuff not captured in the knownResults as the actual TF that the motif comes from might not be known so far.
For mouse/human I typically look at knownResults, but as said, there should not be a dramatic difference, but it is worth checking both.