Search

Accurate modeling of the effects of mutations on protein stability is central to understanding and controlling proteins in myriad natural and applied contexts. Here, we reveal through rigorous quantitative analysis that stability prediction tools often favor mutations that increase stability at the expense of solubility. Moreover, while these tools may accurately identify strongly destabilizing mutations, the experimental effect of mutations predicted to stabilize is actually near neutral on average. The commonly used "classification accuracy" metric obscures this reality; accordingly, we recommend performance measures, such as the Matthews correlation coefficient (MCC). We demonstrate that an absurdly simple machine-learning algorithm-a neural network of just two neurons-unexpectedly achieves high classification accuracy, but its inadequacies are revealed by a low MCC. Despite the above limitations, making multiple mutations markedly improves the prospects for achieving a stabilization target, and modest improvements in the precision of future tools may yield disproportionate gains., Competing Interests: Declaration of Interests The authors declare no competing interests., (Copyright © 2020 Elsevier Ltd. All rights reserved.)

Detailed methods are presented for measurement and study of in vivo mutations and in vitro mutagenesis in human lymphocytes. The methods described include preparation of conditioned medium containing interleukin-2, enumeration of mutant clones, in vitro mutagenesis, and expansion of mutant clones for further study.

The genetic stability of normal and neoplastic lymphocytes was compared by using base-line mutation frequency and mutation rate/cell generation. Mutations at the hypoxanthine-guanine phosphoribosyltransferase locus were studied by enumerating thioguanine-resistant cells in a clonogenic assay. The base-line ("spontaneous") mutation frequency was 1.52 X 10(-6), 6.38 X 10(-6), and 1.06 X 10(-6) for normal cells from three individuals and was 1.16 X 10(-3), 6.08 X 10(-5), and 3.06 X 10(-5) for the three malignant cell lines, Jurkat (JM), HRIK, FMC-Hu1B, respectively. The mutation cell/generation rate was 24.6 X 10(-8), 15 X 10(-8), and 5.5 X 10(-8) for lymphocytes from the three normal individuals, and 666.4 X 10(-8), 52.8 X 10(-8), and 131 X 10(-8) for the three malignant cell lines. The results suggest that neoplastic lymphocytes are more genetically unstable than normal lymphocytes.

Several theories of ageing predict that somatic mutations should increase with age. This prediction was tested for human lymphocytes using a recently developed clonal technique for enumeration of mutations, and an increase of 1.6% per year in mutations with age was detected.