One major focus was having cross-species functionality, so the tool uses Ensembl IDs and can make predictions in 317 different species, including all model organisms. Looking forward (hoping) to seeing how the community uses this to advance transcriptional biology research!

This article presents a tool that integrates multiple data types to improve the accuracy of transcription factor binding site (TFBS) prediction.

Why it matters:
Our results strengthen the idea that subtle tuning of gene activity helped sculpt human-specific cognitive abilities, along with the neuropsychiatric risks that hitchhike on that evolutionary journey.

- In an autism dataset 15 of these TFs and 16 of their target genes are differentially expressed in autism vs. control, most commonly upregulated in developing neurons and downregulated in SV2C- and somatostatin-expressing interneurons, oligodendrocytes.

- The target genes cluster around pathways linked to neurodevelopment (including autism-related processes), craniofacial and connective-tissue biology, tracheal and prostate function, and vision.

What we found:
- Using our tool TFBSFootprinter, we screened the non-coding genome and uncovered 110 transcription factors (TFs) whose binding differs, potentially altering expression of 75 target genes.
- The TFs are enriched for vision, motor neurons, homeobox, and brain.

Background:
- Protein-coding changes between modern humans and Neanderthals are surprisingly scarce—just 96 sites. Regulatory changes are far richer.

Thanks for sharing. I think that it is underutilized for study outside of single cell analyses. In one of our recent studies we used it to correlate cells based on various phenotype measurements after FACS.

If it’s a pre-print somewhere I’d be interested to see your article, or share it after pub!

Are you sure it’s your first, or maybe just the worst?? 🙈

Side note: Greg Bear wrote a great sci-fi series which incorporates Planet X and is the first fiction I’ve ever read which incorporates a more comprehensive view of the extrasolar bodies of the solar system.

3/3
Code for running the analysis and generating the interactive plot here:
github.com/thirtysix/po...

2/3
I grabbed a starting set of 16,765 ancient and modern humans found at AARD (reich.hms.harvard.edu/allen-ancien...), compiled by the David Reich lab, processed the data with ADMIXTOOLS2 and PLINK, and then created an interactive visualization using Python and Bokeh.