Find Genes by Predicted Transcription Factor Targets

Transcription Factor
Type three or more characters in the above textbox to get inline suggestions.
Enter the EntrezGene Identifier or Official Gene Symbol of your transcription factor of interest. After typing 3 characters a list of suggestions will be provided. Choose the term that best matches.
Enter a value between 0.7 and 1. The higher stringency, the greater the scoring threshold and the more likely the predicted site is a true binding site. Higher stringency values will return fewer results. In ErythronDB, only predicted sites associated with a stringencyt of more than 0.7 are considered as potential binding sites.
Give this search a weight
Give this search a name

Search Description

Identify genes that have been computationally predicted to be common transcription factor targets.

Targets were identified by matching consensus sequences and partial-weight matrices (PWMS) to sequences within +/- 1000 bp of the transcriptional start site (TSS) of all genes significantly expressed in the Erythron Database. Matches were identified using the Transcription Element Search System (TESS).

PWMS were obtained from the JASPAR and TransFAC databases.

For each PWM we determined a scoring threshold that conrolled the false positive rate to a workable maximum.

Each potential match was scored using a standard log-odds ratio with a pseudocount of 0.25 added to the observations in the PWMs. Any position in genomics regions of interest that scored at or above the scoring threshold for a PWM and associated with a stringency (true positive rate; the likelihood that a predicted site is a true binding site) greater than 70% was labeled as a predicted binding site.