dms-viz: Structure-informed visualizations for deep mutational scanning and other mutation-based datasets

Abstract

Summary and PurposeUnderstanding how mutations impact a proteins functions is valuable for many types of biological questions. High-throughput techniques such as deep-mutational scanning (DMS) have greatly expanded the number of mutation-function datasets. For instance, DMS has been used to determine how mutations to viral proteins affect antibody escape (Dadonaite et al. 2023), receptor affinity (Starr et al. 2020), and essential functions such as viral genome transcription and replication (Li et al. 2023). With the growth of sequence databases, in some cases the effects of mutations can also be inferred from phylogenies of natural sequences (Bloom and Neher 2023) (Figure 1). O_FIG O_LINKSMALLFIG WIDTH=200 HEIGHT=147 SRC="FIGDIR/small/564578v1_fig1.gif" ALT="Figure 1"> View larger version (31K): org.highwire.dtl.DTLVardef@6c623forg.highwire.dtl.DTLVardef@1a62199org.highwire.dtl.DTLVardef@1e7d71eorg.highwire.dtl.DTLVardef@1dc1252_HPS_FORMAT_FIGEXP M_FIG O_FLOATNOFigure 1.C_FLOATNO Large mutation-associated datasets are used in a variety of experimental contexts. They can be used to map antibody footprints on viral glycoproteins, assess the impact of mutations on protein function in a laboratory setting, and identify patterns of selection from natural mutation frequencies. C_FIG The mutation-based data generated by these approaches is often best understood in the context of a proteins 3D structure; for instance, to assess questions like how mutations that affect antibody escape relate to the physical antibody binding epitope on the protein. However, current approaches for visualizing mutation data in the context of a proteins structure are often cumbersome and require multiple steps and softwares. To streamline the visualization of mutation-associated data in the context of a protein structure, we developed a web-based tool, dms-viz. With dms-viz, users can straightforwardly visualize mutation-based data such as those from DMS experiments in the context of a 3D protein model in an interactive format. See https://dms-viz.github.io/ to use dms-viz.