Similar(60)
No motif was significantly enriched in the JUND conservation-free promoter sets.
Overall, the conservation-free set consistently had the highest recall rate, and its inferred targets were used for all subsequent experiments.
We refer to the ARACNe-inferred promoter set as the conservation-free set because it is assembled without regard to cross-species conservation.
In order to compare OmniMiner on combined-conservation promoters to GibbsModule, we ran GibbsModule on the conservation-free set with the orthologous promoters as additional input.
In our experiments, ARACNe (conservation-free set) significantly outperformed co-expression (p<0.05, by FET), and more narrowly outperformed co-expression*.
Specifically, we ran DME [11] on both the conservation-free and the combined-conservation sets and recorded p-values for DME-identified motifs, reporting motifs with p<0.05 (see Motif evaluation and discovery in Materials and methods).
Finally, members of a third cluster, including JUND, ETS1, ZNF42, SMAD2, LEF1, TAL1, FOXC1, TGIF, and SMAD1, were correctly classified with the help of cross-species conservation but not in the original conservation-free promoter sets.
We obtained 1500 bp promoters for each target gene by selecting [−1000, 500] from Refseq transcription start site locations, eliminating intersecting promoters arbitrarily; we refer to these as the conservation-free sets.
Finally, we merged motif enrichment results independently produced by the conservation-free and the combined-conservation sets, re-ranking motifs according to the best classification relative-error rate achieved in either test (see Figure 1).
Strictly matching significant motifs among the top 5 motifs per TF were recovered for 2/38, 13/38 and 10/38 of the TFs on the conservation-free, combined-conservation, and the combination of the two, respectively.
Following the same procedure described for TRANSFAC motifs, we re-ranked significant motifs based on the best classification relative error rate achieved on either the conservation-free or the combined-conservation sets.
Write better and faster with AI suggestions while staying true to your unique style.
Since I tried Ludwig back in 2017, I have been constantly using it in both editing and translation. Ever since, I suggest it to my translators at ProSciEditing.

Justyna Jupowicz-Kozak
CEO of Professional Science Editing for Scientists @ prosciediting.com