Kaessmann Lab

Half a century ago, King & Wilson proposed that humans differ from chimpanzees mainly through changes in gene regulation. In our new preprint, we used AlphaGenome to map such changes genome-wide, across tissues of the human body! www.biorxiv.org/content/10.6...

Explore all predictions – any gene, any tissue, any substitution – and their combined effects on gene expression in our interactive web application: apps.kaessmannlab.org/human_specific_cre

Half a century ago, King & Wilson proposed that humans differ from chimpanzees mainly through changes in gene regulation. In our new preprint, we used AlphaGenome to map such changes genome-wide, across tissues of the human body! www.biorxiv.org/content/10.6...

When did these changes happen? Using all seven high-coverage Neandertal and Denisovan genomes, we found that nearly all (99.6%) gains now fixed in humans predate the split from archaic hominins, yet a few modern-human-specific gains (e.g. at IRF8 in liver) substantially upregulate their targets.

Half a century ago, King & Wilson proposed that humans differ from chimpanzees mainly through changes in gene regulation. In our new preprint, we used AlphaGenome to map such changes genome-wide, across tissues of the human body! www.biorxiv.org/content/10.6...

Do they change gene expression? Linking gains to target genes (ENCODE-rE2G; Gschwind et al., 2026), we found that genes with more cis-regulatory gains were consistently more likely to be upregulated in humans across all examined cross-species transcriptome datasets.

Half a century ago, King & Wilson proposed that humans differ from chimpanzees mainly through changes in gene regulation. In our new preprint, we used AlphaGenome to map such changes genome-wide, across tissues of the human body! www.biorxiv.org/content/10.6...

How did these gains arise? Mostly by repurposing rather than invention: ~70% occurred in cis-regulatory elements already active in other tissues, where motifs for broadly active transcription factors (e.g. AP-1) provide a permissive evolutionary starting point for gaining activity in a new tissue.

Half a century ago, King & Wilson proposed that humans differ from chimpanzees mainly through changes in gene regulation. In our new preprint, we used AlphaGenome to map such changes genome-wide, across tissues of the human body! www.biorxiv.org/content/10.6...

Does this work? We benchmarked the predictions against measured chromatin accessibility differences between humans and other great apes. AlphaGenome outperformed all other models tested and accurately recovered experimentally measured human-specific changes.

Half a century ago, King & Wilson proposed that humans differ from chimpanzees mainly through changes in gene regulation. In our new preprint, we used AlphaGenome to map such changes genome-wide, across tissues of the human body! www.biorxiv.org/content/10.6...

Millions of non-coding changes arose on the human lineage. Which of them altered gene regulation? In which tissues? With what effect, and when? Functional data are scarce for great apes, unobtainable for extinct hominins – and causal substitutions are hard to pinpoint among the many with no effect.

Half a century ago, King & Wilson proposed that humans differ from chimpanzees mainly through changes in gene regulation. In our new preprint, we used AlphaGenome to map such changes genome-wide, across tissues of the human body! www.biorxiv.org/content/10.6...

Did transposable elements shape brain evolution — and if so, which ones, and in which cell states and lineages? Led by @tyamadat.bsky.social, we explored this question in cerebellum development using sequence-based deep learning models! www.biorxiv.org/content/10.1...

Species-specific accessible HERVL copies contribute to divergent gene expression patterns – including elevated expression of neurodevelopmentally relevant genes such as C9orf72 and MGST2 in humans.

Did transposable elements shape brain evolution — and if so, which ones, and in which cell states and lineages? Led by @tyamadat.bsky.social, we explored this question in cerebellum development using sequence-based deep learning models! www.biorxiv.org/content/10.1...

Accessibility of individual HERVL copies is governed by two complementary factors: the preservation of transcription factor binding motifs and positioning within active chromatin neighborhoods.

Did transposable elements shape brain evolution — and if so, which ones, and in which cell states and lineages? Led by @tyamadat.bsky.social, we explored this question in cerebellum development using sequence-based deep learning models! www.biorxiv.org/content/10.1...

We discovered that HERVLs – primate-specific retrotransposons – are preferentially co-opted as regulatory elements in cerebellar granule cells and other rhombic lip-derived neuroblasts in the pons and medulla, facilitated by the presence of sequences that resemble ATOH1 and NFI binding motifs.

Did transposable elements shape brain evolution — and if so, which ones, and in which cell states and lineages? Led by @tyamadat.bsky.social, we explored this question in cerebellum development using sequence-based deep learning models! www.biorxiv.org/content/10.1...

We found that TEs are significantly enriched in species- and cell type-specific elements – particularly in later-developing, less constrained cell types. This pattern is further supported by datasets spanning multiple human organs across development.

Did transposable elements shape brain evolution — and if so, which ones, and in which cell states and lineages? Led by @tyamadat.bsky.social, we explored this question in cerebellum development using sequence-based deep learning models! www.biorxiv.org/content/10.1...

However, their contribution to mammalian organ development at the cellular level has remained largely unexplored, owing to the lack of cell type-resolved datasets and suitable analytical frameworks.

Did transposable elements shape brain evolution — and if so, which ones, and in which cell states and lineages? Led by @tyamadat.bsky.social, we explored this question in cerebellum development using sequence-based deep learning models! www.biorxiv.org/content/10.1...

In that study, we traced evolutionary history of human cis-regulatory elements (CREs) and highlighted many candidate CRE innovations from single-nucleotide substitutions or small insertions and deletions (indels). But what about TEs?

Did transposable elements shape brain evolution — and if so, which ones, and in which cell states and lineages? Led by @tyamadat.bsky.social, we explored this question in cerebellum development using sequence-based deep learning models! www.biorxiv.org/content/10.1...