Is there any software that, given a DNA sequence as input, can predict what protein/ type of protein can bind to the DNA. The sequence of interest is a potential transcriptional regulator, I want to know for sure (or at least be reasonably certain) before I go to the wet lab to run any experiment.
$\begingroup$
$\endgroup$
4
-
1$\begingroup$ try to do PROKKA predict protein. $\endgroup$– ABDELJALIL SENHAJI RACHIKMay 29, 2020 at 15:31
-
$\begingroup$ Given a crystal structure or model of a protein it is incredibly hard predicting the DNA/RNA sequence it binds. If you do not know the protein, as is your case, homology based methods are the only way. And even then you may have a somewhat conserved non-coding region or motif, but nothing known because an analogue binds there. Transcription factors are present at very low levels so there is not a wealth of data to go by unfortunately, so your options are very limited. $\endgroup$– Matteo FerlaMay 29, 2020 at 15:48
-
$\begingroup$ @MatteoFerla Thanks for the clarification. What kind of homology based approaches can I try, are there any open source softwares $\endgroup$– Anthony GuterresMay 30, 2020 at 16:28
-
1$\begingroup$ @AnthonyGuterres, what I meant to confer was that there is nothing as helpful as you had hoped. PROKKA will annotate a genome (if unannotated) and tell you what are transcription factors by homology to others. Some TF have known binding motifs (very organism dependent), so you'd filter for those with homology to one with a known motifs and go from there —unlikely to give you anything. If it is bacteria, especially thermophilic, you can use guilt by association, i.e. TF is likely nearby on a different transcript. Works sometimes with funghi, while animals and plants are a mess. $\endgroup$– Matteo FerlaJun 1, 2020 at 10:09
Add a comment
|