LM-GVP: A Generalizable Deep Learning Framework for Protein Property Prediction from Sequence and Structure

Proteins perform many essential functions in biological systems and can be successfully developed as bio-therapeutics. It is invaluable to be able to predict their properties based on a proposed sequence and structure. In this study, we developed a novel generalizable deep learning framework, LM-GVP, composed of a protein Language Model (LM) and Graph Neural Network (GNN) to leverage information from both 1D amino acid sequences and 3D structures of proteins. Our approach outperformed the state-of-the-art protein LMs on a variety of property prediction tasks including fluorescence, protease stability, and protein functions from Gene Ontology (GO). We also illustrated insights into how a GNN prediction head can guide the protein LM to better leverage structural information. We envision that our deep learning framework will be generalizable to many protein property prediction problems to greatly accelerate protein engineering and drug development.

Solubility of proteins

<p class="ADMETabstracttext">Solubility is a fundamental protein property that has important connotations for therapeutics and use in diagnosis. Solubility of many proteins is low and affect heterologous overexpression of proteins, formulation of products and their stability. Two processes are related to soluble and solid phase relations. Solubility refers to the process where proteins have correctly folded structure, whereas aggregation is related to the formation of fibrils, oligomers or amorphous particles. Both processes are related to some diseases. Amyloid fibril formation is one of the characteristic features in several neurodegenerative diseases, but it is related to many other diseases, including cancers. Severe complex V deficiency and cataract are examples of diseases due to reduced protein solubility. Methods and approaches are described for prediction of protein solubility and aggregation, as well as predictions of consequences of amino acid substitutions. Finally, protein engineering solutions are discussed. Protein solubility can be increased, although such alterations are relatively rare and can lead to trade-off with some other properties. The aggregation prediction methods mainly aim to detect aggregation-prone sequence patches and then making them more soluble. The solubility predictors utilize a wide spectrum of features.</p>

Variation benchmark datasets: update, criteria, quality and applications

Development of new computational methods and testing their performance has to be carried out using experimental data. Only in comparison to existing knowledge can method performance be assessed. For that purpose, benchmark datasets with known and verified outcome are needed. High-quality benchmark datasets are valuable and may be difficult, laborious and time consuming to generate. VariBench and VariSNP are the two existing databases for sharing variation benchmark datasets used mainly for variation interpretation. They have been used for training and benchmarking predictors for various types of variations and their effects. VariBench was updated with 419 new datasets from 109 papers containing altogether 329 014 152 variants; however, there is plenty of redundancy between the datasets. VariBench is freely available at http://structure.bmc.lu.se/VariBench/. The contents of the datasets vary depending on information in the original source. The available datasets have been categorized into 20 groups and subgroups. There are datasets for insertions and deletions, substitutions in coding and non-coding region, structure mapped, synonymous and benign variants. Effect-specific datasets include DNA regulatory elements, RNA splicing, and protein property for aggregation, binding free energy, disorder and stability. Then there are several datasets for molecule-specific and disease-specific applications, as well as one dataset for variation phenotype effects. Variants are often described at three molecular levels (DNA, RNA and protein) and sometimes also at the protein structural level including relevant cross references and variant descriptions. The updated VariBench facilitates development and testing of new methods and comparison of obtained performances to previously published methods. We compared the performance of the pathogenicity/tolerance predictor PON-P2 to several benchmark studies, and show that such comparisons are feasible and useful, however, there may be limitations due to lack of provided details and shared data. Database URL: http://structure.bmc.lu.se/VariBench

Generation and Application of Fluorescent Anti-Human β2-Microglobulin VHHs via Amino Modification

The functionalization of VHHs enables their application in almost every aspect of biomedical inquiry. Amino modification remains a common strategy for protein functionalization, though is considered to be inferior to site-specific methods and cause protein property changes. In this paper, four anti-β2M VHHs were selected and modified on the amino group by NHS-Fluo. The impacts of amino modification on these VHHs were drastically different, and among all th examples, the modified NB-1 maintained the original stability, bioactivity and homogeneity of unmodified NB-1. Specific recognition of VHHs targeting β2M detected by fluorescence imaging explored the possible applications of VHHs. Via this study, we successfully functionalized the anti-β2M VHHs through amino modification and the results are able to instruct the simple and fast functionalization of VHHs in biomedical researches.

Variation Benchmark Datasets: Update, Criteria, Quality and Applications

ABSTRACTDevelopment of new computational methods and testing their performance has to be done on experimental data. Only in comparison to existing knowledge can method performance be assessed. For that purpose, benchmark datasets with known and verified outcome are needed. High-quality benchmark datasets are valuable and may be difficult, laborious and time consuming to generate. VariBench and VariSNP are the two existing databases for sharing variation benchmark datasets. They have been used for training and benchmarking predictors for various types of variations and their effects. There are 419 new datasets from 109 papers containing altogether 329003373 variants; however there is plenty of redundancy between the datasets. VariBench is freely available athttp://structure.bmc.lu.se/VariBench/. The contents of the datasets vary depending on information in the original source. The available datasets have been categorized into 20 groups and subgroups. There are datasets for insertions and deletions, substitutions in coding and non-coding region, structure mapped, synonymous and benign variants. Effect-specific datasets include DNA regulatory elements, RNA splicing, and protein property predictions for aggregation, binding free energy, disorder and stability. Then there are several datasets for molecule-specific and disease-specific applications, as well as one dataset for variation phenotype effects. Variants are often described at three molecular levels (DNA, RNA and protein) and sometimes also at the protein structural level including relevant cross references and variant descriptions. The updated VariBench facilitates development and testing of new methods and comparison of obtained performance to previously published methods. We compared the performance of the pathogenicity/tolerance predictor PON-P2 to several benchmark studies, and showed that such comparisons are feasible and useful, however, there may be limitations due to lack of provided details and shared data.AUTHOR SUMMARYA prediction method performance can only be assessed in comparison to existing knowledge. For that purpose benchmark datasets with known and verified outcome are needed. High-quality benchmark datasets are valuable and may be difficult, laborious and time consuming to generate. We collected variation datasets from literature, website and databases. There are 419 separate new datasets, which however contain plenty of redundancy. VariBench is freely available athttp://structure.bmc.lu.se/VariBench/. There are datasets for insertions and deletions, substitutions in coding and non-coding region, structure mapped, synonymous and benign variants. Effect-specific datasets include DNA regulatory elements, RNA splicing, and protein property predictions for aggregation, binding free energy, disorder and stability. Then there are several datasets for molecule-specific and disease-specific applications, as well as one dataset for variation phenotype effects. The updated VariBench facilitates development and testing of new methods and comparison of obtained performance to previously published methods. We compared the performance of the pathogenicity/tolerance predictor PON-P2 to several benchmark studies and showed that such comparisons are possible and useful when the details of studies and the datasets are shared.

KStable: A Computational Method for Predicting Protein Thermal Stability Changes by K-Star with Regular-mRMR Feature Selection

Thermostability is a protein property that impacts many types of studies, including protein activity enhancement, protein structure determination, and drug development. However, most computational tools designed to predict protein thermostability require tertiary structure data as input. The few tools that are dependent only on the primary structure of a protein to predict its thermostability have one or more of the following problems: a slow execution speed, an inability to make large-scale mutation predictions, and the absence of temperature and pH as input parameters. Therefore, we developed a computational tool, named KStable, that is sequence-based, computationally rapid, and includes temperature and pH values to predict changes in the thermostability of a protein upon the introduction of a mutation at a single site. KStable was trained using basis features and minimal redundancy–maximal relevance (mRMR) features, and 58 classifiers were subsequently tested. To find the representative features, a regular-mRMR method was developed. When KStable was evaluated with an independent test set, it achieved an accuracy of 0.708.

Preparation of “Open/closed” pores of PLGA-microsphere for controlled release of protein drug

Poly(D,L-lactic-co-glycolic acid) has been extensively used as a controlled release carrier for drug delivery due to its good biocompatibility, biodegradability, and mechanical strength. In this study, porous PLGA microspheres were fabricated by an emulsion-solvent evaporation technique using poly ethylene glycol (PEG) as an extractable porogen and loaded with protein (lysozyme) by suspending them in protein solution. For controlled release of protein, porous microspheres containing lysozyme were treated with water-miscible solvents in aqueous phase for production of pore-closed microspheres. The surface morphology of microspheres were investigated using scanning electron microscopy (SEM) for confirmation of its porous microstructure structure. Protein property after release was observed by enzymatic activity assay. The pore-closing process resulted in nonporous microspheres which exhibited sustained release patterns over an extended period.