An important message taken from human genome sequencing projects is that the human population exhibits approximately 99.9% genetic similarity. Variations in the remaining parts of the genome determine our identity, trace our history and reveal our heritage. The precise delineation of phenotypically causal variants plays a key role in providing accurate personalized diagnosis, prognosis, and treatment of inherited diseases. Several computational methods for achieving such delineation have been reported recently. However, their ability to pinpoint potentially deleterious variants is limited by the fact that their mechanisms of prediction do not account for the existence of different categories of variants. Consequently, their output is biased towards the variant categories that are most strongly represented in the variant databases. Moreover, most such methods provide numeric scores but not binary predictions of the deleteriousness of variants or confidence scores that would be more easily understood by users. We have constructed three datasets covering different types of disease-related variants, which were divided across five categories: (i) regulatory, (ii) splicing, (iii) missense, (iv) synonymous, and (v) nonsense variants. These datasets were used to develop category-optimal decision thresholds and to evaluate six tools for variant prioritization: CADD, DANN, FATHMM, FitCons, FunSeq2 and GWAVA. This evaluation revealed some important advantages of the category-based approach. The results obtained with the five best-performing tools were then combined into a consensus score. Additional comparative analyses showed that in the case of missense variations, protein-based predictors perform better than DNA sequence-based predictors. A user-friendly web interface was developed that provides easy access to the five tools’ predictions, and their consensus scores, in a user-understandable format tailored to the specific features of different categories of variations. To enable comprehensive evaluation of variants, the predictions are complemented with annotations from eight databases. The web server is freely available to the community at http://loschmidt.chemi.muni.cz/predictsnp2.

Natural enzymes are delicate biomolecules possessing only marginal thermodynamic stability. Poorly stable, misfolded, and aggregated proteins lead to huge economic losses in the biotechnology and biopharmaceutical industries. Consequently, there is a need to design optimized protein sequences that maximize stability, solubility, and activity over a wide range of temperatures and pH values in buffers of different composition and in the presence of organic cosolvents. This has created great interest in using computational methods to enhance biocatalysts' robustness and solubility. Suitable methods include (i) energy calculations, (ii) machine learning, (iii) phylogenetic analyses, and (iv) combinations of these approaches. We have witnessed impressive progress in the design of stable enzymes over the last two decades, but predictions of protein solubility and expressibility are scarce. Stabilizing mutations can be predicted accurately using available force fields, and the number of sequences available for phylogenetic analyses is growing. In addition, complex computational workflows are being implemented in intuitive web tools, enhancing the quality of protein stability predictions. Conversely, solubility predictors are limited by the lack of robust and balanced experimental data, an inadequate understanding of fundamental principles of protein aggregation, and a dearth of structural information on folding intermediates. Here we summarize recent progress in the development of computational tools for predicting protein stability and solubility, critically assess their strengths and weaknesses, and identify apparent gaps in data and knowledge. We also present perspectives on the computational design of stable and soluble biocatalysts.