Meanwhile, to directly evaluate the effects of miRNAs, we generated some artificial miRNAs by randomly assigning their target genes from the target genes of all miRNAs.
Chromatin immunoprecipitation experiments of TEL-AML1-expressing BA/F3 cells were integrated with gene expression arrays and differential protein expression data (SILAC) to investigate and distinguish early direct target genes from indirect target genes of the chimeric transcription factor (outlined in Figure 2).
PANDA binds to and sequestrates NF-YA transcription factor from target gene promoters to repress gene expression [ 41].
The presence of NELF somehow facilitates the dissociation of heat shock factor from target genes 54.
For each identified miRNA, we obtained its target genes from MicroCosm Targets database (http://www.ebi.ac.uk/enright-srv/microcosm/htdocs/targets/v5/), in which the candidate miRNA-target relationships were mostly predicted by miRanda algorithm [ 19].
Firstly, for each identified differentially expressed miRNA, we got its target genes from MicroCosm Targets database (http://www.ebi.ac.uk/enright-srv/microcosm/htdocs/targets/v5/), in which the candidate miRNA-target relationships were mostly predicted by miRanda algorithm [ 22].
We are thereby able to identify the specific biological attributes which are most useful for separating target genes from non-targets.
To introduce additional prediction errors, we randomly select 5 %, 10 %, 20 %, 30 %, 40d 50% and 50%rom the targenesene set, set their miR-1 binding scores to 0s and assign their original binding scores to an equal number ofromndomly selecthe non-target geneset
By removing these four genes from the target list and resubmitting the query, a very different network appears.
The genes from the target MHC region were cross-checked against all the 1516 experimentally confirmed estrogen-responsive genes in the KBERG database [ 44].
