The paper introduces mCSM, a novel method for predicting the effects of mutations in proteins using graph-based signatures. mCSM encodes distance patterns between atoms to represent the protein residue environment and train predictive models. The method was evaluated on various tasks, including predicting the impact of mutations on protein stability and interactions with other proteins or nucleic acids. Results show that mCSM performs as well as or better than other widely used methods. It successfully predicts both the direction and magnitude of stability changes in proteins, as well as changes in protein-protein and protein-nucleic acid affinities. The method was applied to predict the effects of mutations in the tumor suppressor protein p53, demonstrating its applicability in a challenging disease scenario. mCSM uses a combination of graph-based atom distance patterns and pharmacophore count vectors to represent the residue environment and account for changes in atom types due to mutations. The method was tested on multiple datasets, including those for protein stability, protein-protein affinity, and protein-nucleic acid affinity. mCSM outperformed other methods in these tasks, showing strong correlation with experimental data. The method is available as a web server and has been validated through extensive comparative experiments. The results indicate that mCSM is a reliable and effective tool for predicting the impact of mutations on protein stability and interactions, with potential applications in understanding disease-related mutations.The paper introduces mCSM, a novel method for predicting the effects of mutations in proteins using graph-based signatures. mCSM encodes distance patterns between atoms to represent the protein residue environment and train predictive models. The method was evaluated on various tasks, including predicting the impact of mutations on protein stability and interactions with other proteins or nucleic acids. Results show that mCSM performs as well as or better than other widely used methods. It successfully predicts both the direction and magnitude of stability changes in proteins, as well as changes in protein-protein and protein-nucleic acid affinities. The method was applied to predict the effects of mutations in the tumor suppressor protein p53, demonstrating its applicability in a challenging disease scenario. mCSM uses a combination of graph-based atom distance patterns and pharmacophore count vectors to represent the residue environment and account for changes in atom types due to mutations. The method was tested on multiple datasets, including those for protein stability, protein-protein affinity, and protein-nucleic acid affinity. mCSM outperformed other methods in these tasks, showing strong correlation with experimental data. The method is available as a web server and has been validated through extensive comparative experiments. The results indicate that mCSM is a reliable and effective tool for predicting the impact of mutations on protein stability and interactions, with potential applications in understanding disease-related mutations.