Overall, the library design choices made when building an MPRA can directly affect which variant effects are measured and how large they appear. Functionally important variants may be underestimated or missed depending on where they are placed in the construct. 6/6
We also find that ~1% of variants overlap Pol III promoters in Alu elements, where both the A and B box motifs must be present in the probe for the variant effect to be captured correctly. 5/6
In a subset of cases, cooperative interactions between TFs contribute to position-specific variant effects. 4/6
The main driver of this positional bias is position-dependent TF activity. TFs don't contribute equally to expression from all locations in a sequence, so variants that perturb TF binding inherit this dependence. Closer to the TSS tends to show larger effects, but this is TF-specific. 3/6
MPRAs typically test variants in the center of ~200bp constructs. We shifted variant positions across the probe and found that while the direction of effects is usually preserved, the magnitude can vary substantially depending on placement. 2/6
MPRAs are the gold-standard tool for measuring how DNA sequences drive gene expression and prioritizing variant effects.
In this preprint we asked: does it matter WHERE you place a variant in an MPRA?
Spoiler: yes, and it might lead you to miss disease-causing variants. 1/6
doi.org/10.64898/202...
SOOOO MANY GENOMICS MODELSSSS! 😱 Often unclear which is best since they benchmark differently! In this preprint, we introduce GAME, a new framework that utilizes APIs to enable sustainable, uniform model evaluation so we can see which is actually best for each task. doi.org/10.1101/2025...
