You could also try the method Lisa2 to infer transcription factors (TFs) underlying DEG results (http://lisa.cistrome.org/ ). One advantage of Lisa2 is that it will have a broad range of TFs it can infer, as it uses 1,000's of ChIP-seq profiles from the cistrome database. You can also input the up-regulated and down-regulated genes into two separate lists (recommended approach) and then compare the results.

Alternatively you can have a look at software like lemone (there are other similar tools around as well) that build transcriptional modules based on expression data (== they try to look for the TF that can explain the expression pattern in your data) , if I remember correctly.