A proteome-wide dependency map of protein interaction motifs

Short linear motifs (SLiMs) are the most ubiquitous protein interaction modules in the unstructured regions of the human proteome. Despite their central role in protein function, our understanding of the contribution of SLiMs to cellular homeostasis remains limited. To address this, we designed base editor libraries to precisely mutate all curated SLiMs and a set of computationally predicted instances defined by SLiM-like evolutionary patterns. By targeting 7,293 SLiM containing regions with 80,473 mutations, we define a SLiM dependency map identifying 450 known and 264 predicted SLiMs required for normal cell proliferation. Notably, the vast majority of essential predicted SLiMs belong to novel classes of SLiMs. We also uncover the binding partners of several predicted SLiMs and provide mechanistic insight into disease causing mutations. Our study provides a proteome-wide resource on SLiM essentiality and highlights the presence of numerous uncharacterised essential SLiMs in the human proteome.