Mutagenesis Data Set Research Articles

Deep mutational scanning reveals transmembrane features governing surface expression of the B cell antigen receptor.

B cells surveil the body for foreign matter using their surface-expressed B cell antigen receptor (BCR), a tetrameric complex comprising a membrane-tethered antibody (mIg) that binds antigens and a signaling dimer (CD79AB) that conveys this interaction to the B cell. Recent cryogenic electron microscopy (cryo-EM) structures of IgM and IgG isotype BCRs provide the first complete views of their architecture, revealing that the largest interaction surfaces between the mIg and CD79AB are in their transmembrane domains (TMDs). These structures support decades of biochemical work interrogating the requirements for assembly of a functional BCR and provide the basis for explaining the effects of mutations. Here we report a focused saturating mutagenesis to comprehensively characterize the nature of the interactions in the mIg TMD that are required for BCR surface expression. We examined the effects of 600 single-amino-acid changes simultaneously in a pooled competition assay and quantified their effects by next-generation sequencing. Our deep mutational scanning results reflect a feature-rich TMD sequence, with some positions completely intolerant to mutation and others requiring specific biochemical properties such as charge, polarity or hydrophobicity, emphasizing the high value of saturating mutagenesis over, for example, alanine scanning. The data agree closely with published mutagenesis and the cryo-EM structures, while also highlighting several positions and surfaces that have not previously been characterized or have effects that are difficult to rationalize purely based on structure. This unbiased and complete mutagenesis dataset serves as a reference and framework for informed hypothesis testing, design of therapeutics to regulate BCR surface expression and to annotate patient mutations.

Abstract 6198: TP53 mutational clusters stratify the Li-Fraumeni syndrome spectrum

Abstract Li-Fraumeni syndrome (LFS) is a highly penetrant cancer predisposition disorder caused by germline variants in the TP53 tumor suppressor gene. LFS has recently been redefined as a ‘spectrum’ disorder to reflect the highly variable cancer types with largely unpredictable ages-of-onset and disease severity. The broad functional gradient associated with different TP53 variants is thought to contribute to LFS heterogeneity, although it is still poorly understood and there is an unmet clinical need for risk stratification strategies to improve variant interpretation and patient care. Here, we performed an unsupervised cluster analysis leveraging p53 mutagenesis datasets (Kato et al., 2003 & Giacomelli et al., 2018) that revealed five TP53 mutational clusters with unique mutation patterns, structural features, and functional consequences. Stratifying germline carriers based on the five clusters exposed important clinical characteristics to consider for patient management, such as ages-of-onset, tumor type development, and survival outcomes. In particular, we discovered an osteosarcoma-prone subgroup comprised of monomeric mutant p53 carriers with aggressive cancer phenotypes. We also identified a cluster of TP53 variants found more often in non-cancer and healthy older populations, and carriers of these variants that developed cancer had less “LFS-like” phenotypes including an older age-of-onset and a significantly higher frequency of colorectal cancers. Remarkably, our TP53 variant clustering strategy could also stratify breast cancer survival among germline carriers. This work provides a new framework to delineate the LFS spectrum toward the development of machine learning-based approaches for personalized cancer surveillance plans. Citation Format: Nicholas W. Fischer, Brianne Laverty, Ran Kafri, Kara N. Maxwell, Emma R. Woodward, Christian Kratz, David Malkin. TP53 mutational clusters stratify the Li-Fraumeni syndrome spectrum [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2024; Part 1 (Regular Abstracts); 2024 Apr 5-10; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2024;84(6_Suppl):Abstract nr 6198.

A predictive model of voltage gating of BK channels via physical modeling and machine learning.

Machine learning has made revolutionary impacts in numerous scientific fields in recent years, especially structural biology. Nonetheless, there remain many challenging problems where data scarcity precludes a purely data-driven approach, such as quantitative modeling of mutational effects on the function of a specific protein. An effective approach for overcoming the data scarcity problem is to incorporate physical principles, such as using information generated from molecular modeling and simulations. Here, we focus on the prediction of voltage gating of the big potassium (BK) channels, which play important roles in muscle contraction and neural transmission. BK channels have been implicated in hypertension, epilepsy and stroke, and yet remains a challenging drug target. At present, cryo-EM structures are available for both the open and closed states, and over 500 mutations have been functionally characterized. We first derived a set of physics-based features to quantify the effects of each possible mutation using computational mutagenesis and molecular simulations, and then trained random forest models using these descriptors on the experimental mutagenesis data set. The resulting model recapitulates the existing mutations with RMSE = 30 mV and R = 0.6 on unseen test data. The model can be further applied to identify potential hotspots for designing novel mutant channels with drastically changed voltage gating properties and to identify possible pathway of allosteric regulation. We expect that the model can be further improved by additional experimental validation and iterative refinement.

ECNet is an evolutionary context-integrated deep learning framework for protein engineering

Machine learning has been increasingly used for protein engineering. However, because the general sequence contexts they capture are not specific to the protein being engineered, the accuracy of existing machine learning algorithms is rather limited. Here, we report ECNet (evolutionary context-integrated neural network), a deep-learning algorithm that exploits evolutionary contexts to predict functional fitness for protein engineering. This algorithm integrates local evolutionary context from homologous sequences that explicitly model residue-residue epistasis for the protein of interest with the global evolutionary context that encodes rich semantic and structural features from the enormous protein sequence universe. As such, it enables accurate mapping from sequence to function and provides generalization from low-order mutants to higher-order mutants. We show that ECNet predicts the sequence-function relationship more accurately as compared to existing machine learning algorithms by using ~50 deep mutational scanning and random mutagenesis datasets. Moreover, we used ECNet to guide the engineering of TEM-1 β-lactamase and identified variants with improved ampicillin resistance with high success rates.

Packpred: Predicting the Functional Effect of Missense Mutations.

Predicting the functional consequences of single point mutations has relevance to protein function annotation and to clinical analysis/diagnosis. We developed and tested Packpred that makes use of a multi-body clique statistical potential in combination with a depth-dependent amino acid substitution matrix (FADHM) and positional Shannon entropy to predict the functional consequences of point mutations in proteins. Parameters were trained over a saturation mutagenesis data set of T4-lysozyme (1,966 mutations). The method was tested over another saturation mutagenesis data set (CcdB; 1,534 mutations) and the Missense3D data set (4,099 mutations). The performance of Packpred was compared against those of six other contemporary methods. With MCC values of 0.42, 0.47, and 0.36 on the training and testing data sets, respectively, Packpred outperforms all methods in all data sets, with the exception of marginally underperforming in comparison to FADHM in the CcdB data set. A meta server analysis was performed that chose best performing methods of wild-type amino acids and for wild-type mutant amino acid pairs. This led to an increase in the MCC value of 0.40 and 0.51 for the two meta predictors, respectively, on the Missense3D data set. We conjecture that it is possible to improve accuracy with better meta predictors as among the seven methods compared, at least one method or another is able to correctly predict ∼99% of the data.

Prediction of Function Determining and Buried Residues Through Analysis of Saturation Mutagenesis Datasets.

Mutational scanning can be used to probe effects of large numbers of point mutations on protein function. Positions affected by mutation are primarily at either buried or at exposed residues directly involved in function, hereafter designated as active-site residues. In the absence of prior structural information, it has not been easy to distinguish between these two categories of residues. We curated and analyzed a set of twelve published deep mutational scanning datasets. The analysis revealed differential patterns of mutational sensitivity and substitution preferences at buried and exposed positions. Prediction of buried-sites solely from the mutational sensitivity data was facilitated by incorporating predicted sequence-based accessibility values. For active-site residues we observed mean sensitivity, specificity and accuracy of 61, 90 and 88% respectively. For buried residues the corresponding figures were 59, 90 and 84% while for exposed non active-site residues these were 98, 44 and 82% respectively. We also identified positions which did not follow these general trends and might require further experimental re-validation. This analysis highlights the ability of deep mutational scans to provide important structural and functional insights, even in the absence of three-dimensional structures determined using conventional structure determination techniques, and also discuss some limitations of the methodology.

AlbaTraDIS: Comparative analysis of large datasets from parallel transposon mutagenesis experiments.

Bacteria need to survive in a wide range of environments. Currently, there is an incomplete understanding of the genetic basis for mechanisms underpinning survival in stressful conditions, such as the presence of anti-microbials. Transposon directed insertion-site sequencing (TraDIS) is a powerful tool to identify genes and networks which are involved in survival and fitness under a given condition by simultaneously assaying the fitness of millions of mutants, thereby relating genotype to phenotype and contributing to an understanding of bacterial cell biology. A recent refinement of this approach allows the roles of essential genes in conditional stress survival to be inferred by altering their expression. These advancements combined with the rapidly falling costs of sequencing now allows comparisons between multiple experiments to identify commonalities in stress responses to different conditions. This capacity however poses a new challenge for analysis of multiple data sets in conjunction. To address this analysis need, we have developed ‘AlbaTraDIS’; a software application for rapid large-scale comparative analysis of TraDIS experiments that predicts the impact of transposon insertions on nearby genes. AlbaTraDIS can identify genes which are up or down regulated, or inactivated, between multiple conditions, producing a filtered list of genes for further experimental validation as well as several accompanying data visualisations. We demonstrate the utility of our new approach by applying it to identify genes used by Escherichia coli to survive in a wide range of different concentrations of the biocide Triclosan. AlbaTraDIS identified all well characterised Triclosan resistance genes, including the primary target, fabI. A number of new loci were also implicated in Triclosan resistance and the predicted phenotypes for a selection of these were validated experimentally with results being consistent with predictions. AlbaTraDIS provides a simple and rapid method to analyse multiple transposon mutagenesis data sets allowing this technology to be used at large scale. To our knowledge this is the only tool currently available that can perform these tasks. AlbaTraDIS is written in Python 3 and is available under the open source licence GNU GPL 3 from https://github.com/quadram-institute-bioscience/albatradis.

TLmutation: Predicting the Effects of Mutations Using Transfer Learning.

A reccurring challenge in bioinformatics is predicting the phenotypic consequence of amino acid variation in proteins. With the recent advancements in sequencing techniques, sufficient genomic data has become available to train models that predict the evolutionary statistical energies, but there is still inadequate experimental data to directly predict functional effects. One approach to overcome this data scarcity is to apply transfer learning and train more models with available data sets. In this study, we propose a set of transfer learning algorithms we call TLmutation, which implements a supervised transfer learning algorithm that transfers knowledge from survival data of a protein to a particular function of that protein. This is followed by an unsupervised transfer learning algorithm that extends the knowledge to a homologous protein. We explore the application of our algorithms in three cases. First, we test the supervised transfer on 17 previously published deep mutagenesis data sets to complete and refine missing data points. We further investigate these data sets to identify which mutations build better predictors of variant functions. In the second case, we apply the algorithm to predict higher-order mutations solely from single point mutagenesis data. Finally, we perform the unsupervised transfer learning algorithm to predict mutational effects of homologous proteins from experimental data sets. These algorithms are generalized to transfer knowledge between Markov random field models. We show the benefit of our transfer learning algorithms to utilize informative deep mutational data and provide new insights into protein variant functions. As these algorithms are generalized to transfer knowledge between Markov random field models, we expect these algorithms to be applicable to other disciplines.

Sequence determinants of polyadenylation-mediated regulation.

The cleavage and polyadenylation reaction is a crucial step in transcription termination and pre-mRNA maturation in human cells. Despite extensive research, the encoding of polyadenylation-mediated regulation of gene expression within the DNA sequence is not well understood. Here, we utilized a massively parallel reporter assay to inspect the effect of over 12,000 rationally designed polyadenylation sequences (PASs) on reporter gene expression and cleavage efficiency. We find that the PAS sequence can modulate gene expression by over five orders of magnitude. By using a uniquely designed scanning mutagenesis data set, we gain mechanistic insight into various modes of action by which the cleavage efficiency affects the sensitivity or robustness of the PAS to mutation. Furthermore, we employ motif discovery to identify both known and novel sequence motifs associated with PAS-mediated regulation. By leveraging the large scale of our data, we train a deep learning model for the highly accurate prediction of RNA levels from DNA sequence alone (R = 0.83). Moreover, we devise unique approaches for predicting exact cleavage sites for our reporter constructs and for endogenous transcripts. Taken together, our results expand our understanding of PAS-mediated regulation, and provide an unprecedented resource for analyzing and predicting PAS for regulatory genomics applications.

Protein model discrimination attempts using mutational sensitivity, predicted secondary structure, and model quality information.

Structure prediction methods often generate a large number of models for a target sequence. Even if the correct fold for the target sequence is sampled in this dataset, it is difficult to distinguish it from other decoy structures. An attempt to solve this problem using experimental mutational sensitivity data for the CcdB protein was described previously by exploiting the correlation of residue depth with mutational sensitivity (r ~ 0.6). We now show that such a correlation extends to four other proteins with localized active sites, and for which saturation mutagenesis datasets exist. We also examine whether incorporation of predicted secondary structure information and the DOPE model quality assessment score, in addition to mutational sensitivity, improves the accuracy of model discrimination using a decoy dataset of 163 targets from CASP. Although most CASP models would have been subjected to model quality assessment prior to submission, we find that the DOPE score makes a substantial contribution to the observed improvement. We therefore also applied the approach to CcdB and four other proteins for which reliable experimental mutational data exist and observe that inclusion of experimental mutational data results in a small qualitative improvement in model discrimination relative to that seen with just the DOPE score. This is largely because of our limited ability to quantitatively predict effects of point mutations on in vivo protein activity. Further improvements in the methodology are required to facilitate improved utilization of single mutant data.

Recursive Neural Networks for Density Estimation Over Generalized Random Graphs.

Structured data in the form of labeled graphs (with variable order and topology) may be thought of as the outcomes of a random graph (RG) generating process characterized by an underlying probabilistic law. This paper formalizes the notions of generalized RG (GRG) and probability density function (pdf) for GRGs. Thence, a "universal" learning machine (combining the encoding module of a recursive neural network and a radial basis functions' network) is introduced for estimating the unknown pdf from an unsupervised sample of GRGs. A maximum likelihood training algorithm is presented and constrained so as to ensure that the resulting model satisfies the axioms of probability. Techniques for preventing the model from degenerate solutions are proposed, as well as variants of the algorithm suitable to the tasks of graphs classification and graphs clustering. The major properties of the machine are discussed. The approach is validated empirically through experimental investigations in the estimation of pdfs for synthetic and real-life GRGs, in the classification of images from the Caltech Benchmark data set and molecules from the Mutagenesis data set, and in clustering of images from the LabelMe data set.

The Essential Genome of Escherichia coli K-12.

ABSTRACTTransposon-directed insertion site sequencing (TraDIS) is a high-throughput method coupling transposon mutagenesis with short-fragment DNA sequencing. It is commonly used to identify essential genes. Single gene deletion libraries are considered the gold standard for identifying essential genes. Currently, the TraDIS method has not been benchmarked against such libraries, and therefore, it remains unclear whether the two methodologies are comparable. To address this, a high-density transposon library was constructed in Escherichia coli K-12. Essential genes predicted from sequencing of this library were compared to existing essential gene databases. To decrease false-positive identification of essential genes, statistical data analysis included corrections for both gene length and genome length. Through this analysis, new essential genes and genes previously incorrectly designated essential were identified. We show that manual analysis of TraDIS data reveals novel features that would not have been detected by statistical analysis alone. Examples include short essential regions within genes, orientation-dependent effects, and fine-resolution identification of genome and protein features. Recognition of these insertion profiles in transposon mutagenesis data sets will assist genome annotation of less well characterized genomes and provides new insights into bacterial physiology and biochemistry.

Quantitative Missense Variant Effect Prediction Using Large-Scale Mutagenesis Data.

Large datasets describing the quantitative effects of mutations on protein function are becoming increasingly available. Here, we leverage these datasets to develop Envision, which predicts the magnitude of a missense variant's molecular effect. Envision combines 21,026 variant effect measurements from nine large-scale experimental mutagenesis datasets, a hitherto untapped training resource, with a supervised, stochastic gradient boosting learning algorithm. Envision outperforms other missense variant effect predictors both on large-scale mutagenesis data and on an independent test dataset comprising 2,312 TP53 variants whose effects were measured using a low-throughput approach. This dataset was never used for hyperparameter tuning or model training and thus serves as an independent validation set. Envision prediction accuracy is also more consistent across amino acids than other predictors. Finally, we demonstrate that Envision's performance improves as more large-scale mutagenesis data are incorporated. We precompute Envision predictions for every possible single amino acid variant in human, mouse, frog, zebrafish, fruit fly, worm, and yeast proteomes (https://envision.gs.washington.edu/).

Analysis of Large-Scale Mutagenesis Data To Assess the Impact of Single Amino Acid Substitutions.

Mutagenesis is a widely used method for identifying protein positions that are important for function or ligand binding. Advances in high-throughput DNA sequencing and mutagenesis techniques have enabled measurement of the effects of nearly all possible amino acid substitutions in many proteins. The resulting large-scale mutagenesis data sets offer a unique opportunity to draw general conclusions about the effects of different amino acid substitutions. Thus, we analyzed 34,373 mutations in 14 proteins whose effects were measured using large-scale mutagenesis approaches. Methionine was the most tolerated substitution, while proline was the least tolerated. We found that several substitutions, including histidine and asparagine, best recapitulated the effects of other substitutions, even when the identity of the wild-type amino acid was considered. The effects of histidine and asparagine substitutions also correlated best with the effects of other substitutions in different structural contexts. Furthermore, highly disruptive substitutions like aspartic and glutamic acid had the most discriminatory power for detecting ligand interface positions. Our work highlights the utility of large-scale mutagenesis data, and our conclusions can help guide future single substitution mutational scans.

Measuring the activity of protein variants on a large scale using deep mutational scanning

Deep mutational scanning marries selection for protein function to high-throughput DNA sequencing in order to quantify the activity of variants of a protein on a massive scale. First, an appropriate selection system for the protein function of interest is identified and validated. Second, a library of variants is created, introduced into the selection system and subjected to selection. Third, library DNA is recovered throughout the selection and deep-sequenced. Finally, a functional score for each variant is calculated on the basis of the change in the frequency of the variant during the selection. This protocol describes the steps that must be carried out to generate a large-scale mutagenesis data set consisting of functional scores for up to hundreds of thousands of variants of a protein of interest. Establishing an assay, generating a library of variants and carrying out a selection and its accompanying sequencing takes on the order of 4-6 weeks; the initial data analysis can be completed in 1 week.

A Novel Test for Selection on cis-Regulatory Elements Reveals Positive and Negative Selection Acting on Mammalian Transcriptional Enhancers

Measuring natural selection on genomic elements involved in the cis-regulation of gene expression--such as transcriptional enhancers and promoters--is critical for understanding the evolution of genomes, yet it remains a major challenge. Many studies have attempted to detect positive or negative selection in these noncoding elements by searching for those with the fastest or slowest rates of evolution, but this can be problematic. Here, we introduce a new approach to this issue, and demonstrate its utility on three mammalian transcriptional enhancers. Using results from saturation mutagenesis studies of these enhancers, we classified all possible point mutations as upregulating, downregulating, or silent, and determined which of these mutations have occurred on each branch of a phylogeny. Applying a framework analogous to Ka/Ks in protein-coding genes, we measured the strength of selection on upregulating and downregulating mutations, in specific branches as well as entire phylogenies. We discovered distinct modes of selection acting on different enhancers: although all three have experienced negative selection against downregulating mutations, the selection pressures on upregulating mutations vary. In one case, we detected positive selection for upregulation, whereas the other two had no detectable selection on upregulating mutations. Our methodology is applicable to the growing number of saturation mutagenesis data sets, and provides a detailed picture of the mode and strength of natural selection acting on cis-regulatory elements.

Classification of graphical data made easy

The classification of graphical patterns (i.e., data that are represented in the form of labeled graphs) is a problem that has been receiving considerable attention by the machine learning community in recent years. Solutions to the problem would be valuable to a number of applications, ranging from bioinformatics and cheminformatics to Web-related tasks, structural pattern recognition for image processing, etc. Several approaches have been proposed so far, e.g. inductive logic programming and kernels for graphs. Connectionist models were introduced too, namely recursive neural nets (RNN) and graph neural nets (GNN). Although their theoretical properties are sound and thoroughly understood, RNNs and GNNs suffer some drawbacks that may limit their application. This paper introduces an alternative connectionist framework for learning discriminant functions over graphical data. The approach is simple and suitable to maximum-a-posteriori classification of broad families of graphs, and overcomes some limitations of RNNs and GNNs. The idea is to describe a graph as an algebraic relation, i.e. as a subset of the Cartesian product. The class-posterior probabilities given the relation are then reduced (under an iid assumption) to products of probabilistic quantities, estimated using a multilayer perceptron. Empirical evidence shows that, in spite of its simplicity, the technique compares favorably with established approaches on several tasks involving different graphical representations of the data. In particular, in the classification of molecules from the Mutagenesis dataset ( friendly + unfriendly ) the best result to date (93.91%) is obtained.

Discretization Numerical Data for Relational Data with One-to-Many Relations

Problem statement: Handling numerical data stored in a relational data base has been performed differently from handling those numerical data stored in a single table due to the multiple occurrences (one-to-many association) of an individ ual record in the non-target table and non-determin ate relations between tables. Numbers in Multi-Relation al Data Mining (MRDM) were often discretized, after considering the schema of the relational data base. Study the effects of taking the one-to-many association issue into consideration in the process of discretizing continuous numbers. Approach: Different alternatives for dealing with continuous attributes in MRDM were considered in this study, namely equal-width (EWD), Equal-Height (EH), equal-weight (EWG) and Entropy-Based (EB). The discretization procedures considered in this study included algorithms that were not depended on the multi-relational structure of the data and also tha t are sensitive to this structure. A new method of discretization, called the entropy instance-based ( EIB) discretization method was implemented and evaluated with respect to C4.5 on the two well-known multi-relational databases that include the Mutagenesis dataset and the Hepatitis dataset for D iscovery Challenge PKDD 2005. Results: When the number of bins, b, is big (b = 8), the entropy- instance-based discretization method produced bette r data summarization results compared to the other di scretization methods, in the mutagenesis dataset. I n contrast, for the hepatitis dataset, the entropy-in stance-based discretization method produced better data summarization results for all values of b, com pared to the other discretization methods. In the Hepatitis dataset, all discretization methods produ ced higher average performance accuracy (%) for partitional clustering technique, compared to the h ierarchical technique. Conclusion: These results demonstrated that entropy-based discretization can be improved by taking into consideration the multiple-instance problem. It was also found that t he partitional clustering technique produced better performance accuracy compared to the one produced by hierarchical clustering technique.

Comparison of graph-based and logic-based multi-relational data mining

We perform an experimental comparison of the graph-based multi-relational data mining system, Subdue, and the inductive logic programming system, CProgol, on the Mutagenesis dataset and various artificially generated Bongard problems. Experimental results indicate that Subdue can significantly outperform CProgol while discovering structurally large multi-relational concepts. It is also observed that CProgol is better at learning semantically complicated concepts and it tends to use background knowledge more effectively than Subdue. An analysis of the results indicates that the differences in the performance of the systems are a result of the difference in the expressiveness of the logic-based and the graph-based representations. The ability of graph-based systems to learn structurally large concepts comes from the use of a weaker representation whose expressiveness is intermediate between propositional and first-order logic. The use of this weaker representation is advantageous while learning structurally large concepts but it limits the learning of semantically complicated concepts and the utilization background knowledge.