Question

TF-gene database

0

Entering edit mode

2.7 years ago

Alessiac • 0

Hi everyone, I'm currently writing my thesis and I need to find the TF that interact with a series of gene. I have tried to use different database such as TF2DNA, hTF, TRRUST and TcoF but each one gave me different set of TF for the same gene. Now, i'm a beginner but this has left me confused, how do I find the "right one"?

gene database TF • 1.1k views

ADD COMMENT • link 2.7 years ago by Alessiac • 0

0

Entering edit mode

More databases in this post: Transcription factors database

ADD REPLY • link 2.7 years ago by Hamid Ghaedi 3.3k

0

Entering edit mode

Thank you, I'll sure take a look. But my question still is: why every database give me different result? it's not just one different TF or two, it's the whole set. how do I have to use this different data?

ADD REPLY • link 2.7 years ago by Alessiac • 0

1

Entering edit mode

You might need to read more about each database and see what kind of data each database provides you with. For example, JASPAR :

JASPAR is a regularly maintained open-access database storing manually curated transcription factors (TF) binding profiles as position frequency matrices (PFMs)

While others like GRNdb is presenting predicted results:

GRNdb is a freely accessible and user-friendly database for conveniently exploring and visualizing the predicted regulatory networks formed by transcription factors (TFs) and downstream target genes (termed regulons) based on large-scale RNA-seq data as well as the known TF-target relationships for various human and mouse conditions.

Other than that, even if two databases presented the same data type (experimental vs computational), the result of a given query on each of them wouldn't necessarily be the same. This might be true if databases use different sources of data, using different algorithms to conclude on gene-TF relationship and ... Generally speaking, you have to choose which one is more appropriate for your purpose. If your study design requires you to stick with low false positives, then go with databases that provide data from strictly designed experiments. If you want to be more comprehensive at expense of taking more false-positive calls then databases with predictive results may be more helpful.