Large-scale in-silico statistical mutagenesis analysis sheds light on the deleteriousness landscape of the human proteome

Abstract

Next generation sequencing technologies are providing increasing amounts of sequencing data, paving the way for improvements in clinical genetics and precision medicine. The interpretation of the observed genomic variants in the light of their phenotypic effects is thus emerging as a crucial task to solve in order to advance our understanding of how exomic variants affect proteins and how the proteins’ functional changes affect human health. Since the experimental evaluation of the effects of every observed variant is unfeasible, Bioinformatics methods are being developed to address this challenge in-silico, by predicting the impact of millions of variants, thus providing insight into the deleteriousness landscape of entire proteomes. Here we show the feasibility of this approach by using the recently developed DEOGEN2 variant-effect predictor to perform the largest in-silico mutagenesis scan to date. We computed the deleteriousness score of 170 million variants over 15000 human proteins and we analysed the results, investigating how the predicted deleteriousness landscape of the proteins relates to known functionally and structurally relevant protein regions and biophysical properties. Moreover, we qualitatively validated our results by comparing them with two mutagenesis studies targeting two specific proteins, showing the consistency of DEOGEN2 predictions with respect to experimental data.

DNA sequences performs as natural language processing by exploiting deep learning algorithm for the identification of N4-methylcytosine

“…2. The first method to decipher a convolution neural network model for computational and statistical biology is silico mutagenesis which was used in various scientific works 29,52 . It is operated by mutating each nucleotide by a single base of sequences with a fixed length of four nucleotide A, C, G, T. In this systematical methodology, the model restore each outcome of resulted mutation and keeps the output as an absolute difference.…”

“…We used a K-fold crossvalidation method for different values of K. Then, we applied five evaluation metrics to assess the model. We also employed two applications, silico mutagenesis 29 and saliency 30 map to interpret the predictive deep learning model and influence of important features. The results of the predictive model showed outperformance when compared to the state-of-the-art model.…”

Fast and exact single and double mutation-response scanning of proteins

“…These methods should be useful for a wide range of potential applications, such as predicting evolutionary divergence of protein structures, 16,17 detecting and interpreting pathological mutations, 14,21,22 and detecting compensating mutations and rescue sites. 15 To finish, I mention two possible lines of further development.…”

DNC4mC-Deep: Identification and Analysis of DNA N4-Methylcytosine Sites Based on Different Encoding Schemes By Using Deep Learning

“…The first most authenticated and reliable method to interpret a CNN model for computational biology is silico mutagenesis which is used in several research works [35,63,64]. We computationally mutated the nucleotides by mutating each nucleotide of a single sequence with a fixed length of five nucleotides A, C, G, T, and N. During this systematic approach, the model recomposes the output of every mutation and stores the output as an absolute difference.…”

“…To evaluate and analyze the results of the model on each encoding scheme, we used the performance evaluation metrics. We also presented two different applications; the first one is silico mutagenesis [35] representation using heat maps, and the second is distinguishing the most significant portions of a sequence using saliency maps [36]. After getting the results from the model by all six different feature encoding methods, we ended up with that Dinucleotide composition (DNC) is outperforms from all six encoding schemes and the state-of-the-art model.…”