Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “proteins binding”

Search indexed NASA NTRS and DOE OSTI research on propulsion, heat transfer, battery materials and energy systems. Follow report and document links to the original sources.

Quote a phrase for an exact phrase match. Source license links do not imply unrestricted reuse.

At least 163 records · Page 9
All sourcesNASA NTRSDOE OSTI
DOE OSTI2025

CAML: Commutative Algebra Machine Learning─A Case Study on Protein–Ligand Binding Affinity Prediction

Recently, Suwayyid and Wei introduced commutative algebra as an emerging paradigm for machine learning and data science. In this work, we propose commutative algebra machine learning (CAML) for the prediction of protein−ligand binding affinities. Specifically, we apply persistent Stanley−Reisner theory, a key concept in combinatorial commutative algebra, to the affinity predictions of protein−ligand binding and metalloprotein−ligand binding. We present three new algorithms, i.e., element-specific commutative algebra, category-specific commutative algebra, and commutative algebra on bipartite complexes, to tackle the complexity of data involved in (metallo) protein−ligand complexes. We show that the proposed CAML outperforms other state-of-theart methods in (metallo) protein−ligand binding affinity predictions, indicating the great potential of commutative algebra learning.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗
NASA NTRS1994

Growth Factors and Tension-Induced Skeletal Muscle Growth

The project investigated biochemical mechanisms to enhance skeletal muscle growth, and developed a computer based mechanical cell stimulator system. The biochemicals investigated in this study were insulin/(Insulin like Growth Factor) IGF-1 and Steroids. In order to analyze which growth factors are essential for stretch-induced muscle growth in vitro, we developed a defined, serum-free medium in which the differentiated, cultured avian muscle fibers could be maintained for extended periods of time. The defined medium (muscle maintenance medium, MM medium) maintains the nitrogen balance of the myofibers for 3 to 7 days, based on myofiber diameter measurements and myosin heavy chain content. Insulin and IGF-1, but not IGF-2, induced pronounced myofiber hypertrophy when added to this medium. In 5 to 7 days, muscle fiber diameters increase by 71 % to 98% compared to untreated controls. Mechanical stimulation of the avian muscle fibers in MM medium increased the sensitivity of the cells to insulin and IGF-1, based on a leftward shift of the insulin dose/response curve for protein synthesis rates. (54). We developed a ligand binding assay for IGF-1 binding proteins and found that the avian skeletal muscle cultures produced three major species of 31, 36 and 43 kD molecular weight (54) Stretch of the myofibers was found to have no significant effect on the efflux of IGF-1 binding proteins, but addition of exogenous collagen stimulated IGF-1 binding protein production 1.5 to 5 fold. Steroid hormones have a profound effect on muscle protein turnover rates in vivo, with the stress-related glucocorticoids inducing rapid skeletal muscle atrophy while androgenic steroids induce skeletal muscle growth. Exercise in humans and animals reduces the catabolic effects of glucocorticoids and may enhance the anabolic effects of androgenic steroids on skeletal muscle. In our continuing work on the involvement of exogenrus growth factors in stretch-induced avian skeletal muscle growth, we have performed experiments to determine whether mechanical stimulation of cultured avian muscle cells alters their response to anabolic steroids or glucocorticoids. In static cultures, testosterone had no effect on muscle cell growth, but 5alpha-dihydrotestosterone and the synthetic steroid stanozolol increased cell growth by up to 18% and 30%, respectively, after a three day exposure. We completed development of a new IBM-based mechanical cell stimulator system to provide greater flexibility in operating and monitoring our experiments. Our previous long term studies on myofiber growth were designed around a perfusion system of our own design. We have recently changed to performing these studies using a modified CELLCO cartridge bioreactor system Z since it has been certified as the ground-based model for the Shuttle's Space Tissue Loss (STL) F= Cell Culture Module. The current goals of this aspect of the project are three fold: 1) to design a Z cell culture system for studying avian skeletal myofiber atrophy on the Shuttle and Space Station; 0 2) to expand the use of bioreactors to cells which do not grow in either suspension or attached to the hollow fibers; and 3) to combine the bioreactor system with our computerized mechanical cell stimulator to have a better in vitro model to study tension/gravity/stretch regulation of skeletal muscle size. Preliminary studies also reported on involved : (1) how release of tension can induce rapid atrophy of tissues cultured avian skeletal muscle cells, and (2) a mechanism to transfer and maintain avian skeletal muscle organoids in modified cartridges in the Space Tissue Loss Module.

Vandenburgh, Herman H.↗
DOE OSTI2020

Predicting Small Molecule Transfer Free Energies by Combining Molecular Dynamics Simulations and Deep Learning

Accurately predicting small molecule partitioning and hydrophobicity is critical in the drug discovery process. There are many heterogeneous chemical environments within a cell and entire human body. For example, drugs must be able to cross the hydrophobic cellular membrane to reach their intracellular targets, and hydrophobicity is an important driving force for drug–protein binding. Atomistic molecular dynamics (MD) simulations are routinely used to calculate free energies of small molecules binding to proteins, crossing lipid membranes, and solvation but are computationally expensive. Machine learning (ML) and empirical methods are also used throughout drug discovery but rely on experimental data, limiting the domain of applicability. We present atomistic MD simulations calculating 15,000 small molecule free energies of transfer from water to cyclohexane. This large data set is used to train ML models that predict the free energies of transfer. We show that a spatial graph neural network model achieves the highest accuracy, followed closely by a 3D-convolutional neural network, and shallow learning based on the chemical fingerprint is significantly less accurate. A mean absolute error of ~4 kJ/mol compared to the MD calculations was achieved for our best ML model. We also show that including data from the MD simulation improves the predictions, tests the transferability of each model to a diverse set of molecules, and show multitask learning improves the predictions. This work provides insight into the hydrophobicity of small molecules and ML cheminformatics modeling, and our data set will be useful for designing and testing future ML cheminformatics methods.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗
DOE OSTI2021

Modification of cell wall polysaccharide guides cell division in Streptococcus mutans

In ovoid-shaped, Gram-positive bacteria, MapZ guides FtsZ-ring positioning at cell equators. The cell wall of the ovococcus Streptococcus mutans contains peptidoglycan decorated with serotype c carbohydrates (SCCs). In the present study, we identify the major cell separation autolysin AtlA as an SCC-binding protein. AtlA binding to SCC is attenuated by the glycerol phosphate (GroP) modification. Using fluorescently labeled AtlA constructs, we mapped SCC distribution on the streptococcal surface, revealing enrichment of GroP-deficient immature SCCs at the cell poles and equators. The immature SCCs co-localize with MapZ at the equatorial rings throughout the cell cycle. In GroP-deficient mutants, AtlA is mislocalized, resulting in dysregulated cellular autolysis. These mutants display morphological abnormalities associated with MapZ mislocalization, leading to FtsZ-ring misplacement. Altogether, our data support a model in which maturation of a cell wall polysaccharide provides the molecular cues for the recruitment of cell division machinery, ensuring proper daughter cell separation and FtsZ-ring positioning.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗
NASA NTRSDate not supplied

Considering Cell Proliferation to Optimize Detection of Radiation-induced 53BP1+ Foci in 15 Mouse Strains ex vivo

Due to high metabolic activity, proliferating cells continuously generate free radicals, which induce DNA double strand breaks (DSB). Fluorescently tagged nuclear foci of DNA repair protein 53 binding protein-1(53BP1) are used as a standard metric for measuring DSB formation at baseline and in response to environmental insults such as radiation. Here we demonstrate that the background level of spontaneous 53BP1+ foci formation can be modeled mathematically as a function of cell confluence, which is a metric of proliferation rate. This model was validated using spontaneous 53BP1+ foci data from 68 cultures of primary skin fibroblasts derived from 15 different strains of mice, showing a ~10 fold decrease from low to full confluence that is independent of mouse strain. We developed an online open access tool to correct for the impact of cell confluence on the detection of radiation-induced 53BP1+ foci (RIF). This tool provides guidelines for the number of cells required to reach statistical significance for the detection of excess foci induced by low doses of ionizing radiation as a function of confluence and time post-irradiation. We hope that this quantification tool will help the radiation biology community in the design of future experiments that utilize 53BP1+ foci-based quantification of radiation responses in vitro. Our “tool for enhanced results of RIF in cells” (terRIFic) can be found at: https://radbiolab.shinyapps.io/terrific/

Radiation, DNA damage, confluence↗
DOE OSTI2022

Large Scale Study of Ligand–Protein Relative Binding Free Energy Calculations: Actionable Predictions from Statistically Robust Protocols

The accurate and reliable prediction of protein–ligand binding affinities can play a central role in the drug discovery process as well as in personalized medicine. Of considerable importance during lead optimization are the alchemical free energy methods that furnish an estimation of relative binding free energies (RBFE) of similar molecules. Recent advances in these methods have increased their speed, accuracy, and precision. This is evident from the increasing number of retrospective as well as prospective studies employing them. However, such methods still have limited applicability in real-world scenarios due to a number of important yet unresolved issues. Here, we report the findings from a large data set comprising over 500 ligand transformations spanning over 300 ligands binding to a diverse set of 14 different protein targets which furnish statistically robust results on the accuracy, precision, and reproducibility of RBFE calculations. We use ensemble-based methods which are the only way to provide reliable uncertainty quantification given that the underlying molecular dynamics is chaotic. These are implemented using TIES (Thermodynamic Integration with Enhanced Sampling). Results achieve chemical accuracy in all cases. Ensemble simulations also furnish information on the statistical distributions of the free energy calculations which exhibit non-normal behavior. We find that the “enhanced sampling” method known as replica exchange with solute tempering degrades RBFE predictions. We also report definitively on numerous associated alchemical factors including the choice of ligand charge method, flexibility in ligand structure, and the size of the alchemical region including the number of atoms involved in transforming one ligand into another. Our findings provide a key set of recommendations that should be adopted for the reliable application of RBFE methods.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗
DOE OSTI2022

Protein–Ligand Binding Free-Energy Calculations with ARROW–A Purely First-Principles Parameterized Polarizable Force Field

Protein–ligand binding free-energy calculations using molecular dynamics (MD) simulations have emerged as a powerful tool for in silico drug design. Here, we present results obtained with the ARROW force field (FF)–a multipolar polarizable and physics-based model with all parameters fitted entirely to high-level ab initio quantum mechanical (QM) calculations. ARROW has already proven its ability to determine solvation free energy of arbitrary neutral compounds with unprecedented accuracy. The ARROW FF parameterization is now extended to include coverage of all amino acids including charged groups, allowing molecular simulations of a series of protein–ligand systems and prediction of their relative binding free energies. We ensure adequate sampling by applying a novel technique that is based on coupling the Hamiltonian Replica exchange (HREX) with a conformation reservoir generated via potential softening and nonequilibrium MD. ARROW provides predictions with near chemical accuracy (mean absolute error of ~0.5 kcal/mol) for two of the three protein systems studied here (MCL1 and Thrombin). The third protein system (CDK2) reveals the difficulty in accurately describing dimer interaction energies involving polar and charged species. Overall, for all of the three protein systems studied here, ARROW FF predicts relative binding free energies of ligands with a similar accuracy level as leading nonpolarizable force fields.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗
DOE OSTI2020

Assessment of software methods for estimating protein-protein relative binding affinities

A growing number of computational tools have been developed to accurately and rapidly predict the impact of amino acid mutations on protein-protein relative binding affinities. Such tools have many applications, for example, designing new drugs and studying evolutionary mechanisms. In the search for accuracy, many of these methods employ expensive yet rigorous molecular dynamics simulations. By contrast, non-rigorous methods use less exhaustive statistical mechanics, allowing for more efficient calculations. However, it is unclear if such methods retain enough accuracy to replace rigorous methods in binding affinity calculations. This trade-off between accuracy and computational expense makes it difficult to determine the best method for a particular system or study. Here, eight non-rigorous computational methods were assessed using eight antibody-antigen and eight non-antibody-antigen complexes for their ability to accurately predict relative binding affinities (ΔΔG) for 654 single mutations. In addition to assessing accuracy, we analyzed the CPU cost and performance for each method using a variety of physico-chemical structural features. This allowed us to posit scenarios in which each method may be best utilized. Most methods performed worse when applied to antibody-antigen complexes compared to non-antibody-antigen complexes. Rosetta-based JayZ and EasyE methods classified mutations as destabilizing (ΔΔG < -0.5 kcal/mol) with high (83–98%) accuracy and a relatively low computational cost for non-antibody-antigen complexes. Some of the most accurate results for antibody-antigen systems came from combining molecular dynamics with FoldX with a correlation coefficient (r) of 0.46, but this was also the most computationally expensive method. Overall, our results suggest these methods can be used to quickly and accurately predict stabilizing versus destabilizing mutations but are less accurate at predicting actual binding affinities. This study highlights the need for continued development of reliable, accessible, and reproducible methods for predicting binding affinities in antibody-antigen proteins and provides a recipe for using current methods.

59 BASIC BIOLOGICAL SCIENCES↗
DOE OSTI2024

Expression, purification, and characterization of diacylated Lipo-YcjN from Escherichia coli

YcjN is a putative substrate binding protein expressed from a cluster of genes involved in carbohydrate import and metabolism in Escherichia coli. Here, we determine the crystal structure of YcjN to a resolution of 1.95 Å, revealing that its three-dimensional structure is similar to substrate binding proteins in subcluster D-I, which includes the well-characterized maltose binding protein. Furthermore, we found that recombinant overexpression of YcjN results in the formation of a lipidated form of YcjN that is posttranslationally diacylated at cysteine 21. Comparisons of size-exclusion chromatography profiles and dynamic light scattering measurements of lipidated and nonlipidated YcjN proteins suggest that lipidated YcjN aggregates in solution via its lipid moiety. Additionally, bioinformatic analysis indicates that YcjN-like proteins may exist in both Bacteria and Archaea, potentially in both lipidated and nonlipidated forms. Together, our results provide a better understanding of the aggregation properties of recombinantly expressed bacterial lipoproteins in solution and establish a foundation for future studies that aim to elucidate the role of these proteins in bacterial physiology.

Escherichia coli↗
DOE OSTI2025

Protein-ligand binding affinity prediction using multi-instance learning with docking structures

Recent advances in 3D structure-based deep learning approaches demonstrate improved accuracy in predicting protein-ligand binding affinity in drug discovery. These methods complement physics-based computational modeling such as molecular docking for virtual high-throughput screening. Despite recent advances and improved predictive performance, most methods in this category primarily rely on utilizing co-crystal complex structures and experimentally measured binding affinities as both input and output data for model training. Nevertheless, co-crystal complex structures are not readily available and the inaccurate predicted structures from molecular docking can degrade the accuracy of the machine learning methods. We introduce a novel structure-based inference method utilizing multiple molecular docking poses for each complex entity. Our proposed method employs multi-instance learning with an attention network to predict binding affinity from a collection of docking poses. We validate our method using multiple datasets, including PDBbind and compounds targeting the main protease of SARS-CoV-2. The results demonstrate that our method leveraging docking poses is competitive with other state-of-the-art inference models that depend on co-crystal structures. This method offers binding affinity prediction without requiring co-crystal structures, thereby increasing its applicability to protein targets lacking such data.

97 MATHEMATICS AND COMPUTING↗
NASA NTRS1989

Polyamine binding to proteins in oat and Petunia protoplasts

Previous work (A Apelbaum et al. [1988] Plant Physiol 88: 996-998) has demonstrated binding of labeled spermidine (Spd) to a developmentally regulated 18 kilodalton protein in tobacco tissue cultures derived from thin surface layer explants. To assess the general importance of such Spd-protein complexes, we attempted bulk isolation from protoplasts of Petunia and oat (Avena sativa). In Petunia, as in tobacco, fed radioactive Spd is bound to protein, but in oat, Spd is first converted to 1,3,-diaminopropane (DAP), probably by polyamine oxidase action. In oat, binding of DAP to protein depends on age of donor leaf and conditions of illumination and temperature, and the extraction of the DAP-protein complex depends upon buffer and pH. The yield of the DAP-protein complex was maximized by extraction of frozen-thawed protoplasts with a pH 8.8 carbonate buffer containing SDS. Its molecular size, based on Sephacryl column fractionation of ammonium sulfate precipitated material, exceeded 45 kilodaltons. Bound Spd or DAP can be released from their complexes by the action of Pronase, but not DNAse, RNAse, or strong salt solutions, indicating covalent attachment to protein.

NASA Discipline Plant Biology↗
NASA NTRS2002

Patterns of gene expression in atrophying skeletal muscles: response to food deprivation

During fasting and many systemic diseases, muscle undergoes rapid loss of protein and functional capacity. To define the transcriptional changes triggering muscle atrophy and energy conservation in fasting, we used cDNA microarrays to compare mRNAs from muscles of control and food-deprived mice. Expression of >94% of genes did not change, but interesting patterns emerged among genes that were differentially expressed: 1) mRNAs encoding polyubiquitin, ubiquitin extension proteins, and many (but not all) proteasome subunits increased, which presumably contributes to accelerated protein breakdown; 2) a dramatic increase in mRNA for the ubiquitin ligase, atrogin-1, but not most E3s; 3) a significant suppression of mRNA for myosin binding protein H (but not other myofibrillar proteins) and IGF binding protein 5, which may favor cell protein loss; 4) decreases in mRNAs for several glycolytic enzymes and phosphorylase kinase subunits, and dramatic increases in mRNAs for pyruvate dehydrogenase kinase 4 and glutamine synthase, which should promote glucose sparing and gluconeogenesis. During fasting, metallothionein mRNA increased dramatically, mRNAs for extracellular matrix components fell, and mRNAs that may favor cap-independent mRNA translation rose. Significant changes occurred in mRNAs for many growth-related proteins and transcriptional regulators. These transcriptional changes indicate a complex adaptive program that should favor protein degradation and suppress glucose oxidation in muscle. Similar analysis of muscles atrophying for other causes is allowing us to identify a set of atrophy-specific changes in gene expression.

Non-NASA Center↗
DOE OSTI2025

QM Investigation of Rare Earth Ion Interactions with First Hydration Shell Waters and Protein-Based Coordination Models

Here, conventional methods for extracting rare earth metals (REMs) from mined mineral ores are inefficient, expensive, and environmentally damaging. Recent discovery of lanmodulin (LanM), a protein that coordinates REMs with high-affinity and selectivity over competing ions, provides inspiration for new REM refinement methods. Here, we used quantum mechanical (QM) methods to investigate trivalent lanthanide cation (Ln 3+ ) interactions with coordination systems representing bulk solvent water and protein binding sites. Energy decomposition analysis (EDA) showed differences in the energetic components of Ln 3+ interaction with representatives of solvent (water, H 2 O) and protein binding sites (acetate, CH 3 COO – ), highlighting the importance of accurate description of electrostatics and polarization in computational modeling of REM interactions with biological and bioinspired molecules. Relative binding free energies were obtained for Ln 3+ with coordination complexes originating from binding sites in PDB structures of a lanthanum binding peptide (PDB entry 7CCO) and LanM, with explicit consideration of the first hydration shell waters, according to quasi-chemical theory (QCT). Beyond the first shell, the bulk solvent environment was represented with an implicit continuum model. Ln 3+ interactions with (H 2 O) 9 and both binding site models became more favorable, moving down the periodic series. This trend was more pronounced with the protein binding site models than with water, resulting in affinity increasing with periodic number, except for the last REM, Lu 3+ , which bound less favorably than the preceding element, Yb 3+ . Using the truncated 7CCO binding site model, the magnitude and trend of the experimental Ln 3+ relative binding free energies for the whole 7CCO peptide were reproduced. Conversely, the previously reported experimental data for LanM show a preference for the earlier lanthanides; this is likely due to longer-range interactions and cooperative effects, which are not represented by the reduced models. Using the truncated 7CCO binding site model, the magnitude and trend of the experimental Ln 3+ relative binding free energies for the whole 7CCO peptide were reproduced. In contrast to the previously reported experimental data for LanM, the peptide preferentially binds the earlier lanthanides. This difference likely arises due to longer-range interactions and cooperative effects not represented by the peptide. Further investigation of Ln 3+ interactions with whole proteins using polarizable molecular mechanics models with explicit solvent is warranted to understand the influence of longer-ranged interactions, cooperativity, and bulk solvent. Nevertheless, the present work provides new insights into Ln 3+ interactions with biomolecules and presents an effective computational platform for designing specific single-site REM binding peptides more efficiently.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗
DOE OSTI2024

Bacterial hemophilin homologs and their specific type eleven secretor proteins have conserved roles in heme capture and are diversifying as a family

Cellular life relies on enzymes that require metals, which must be acquired from extracellular sources. Bacteria utilize surface and secreted proteins to acquire such valuable nutrients from their environment. These include the cargo proteins of the type eleven secretion system (T11SS), which have been connected to host specificity, metal homeostasis, and nutritional immunity evasion. This Sec-dependent, Gram-negative secretion system is encoded by organisms throughout the phylum Proteobacteria, including human pathogens Neisseria meningitidis, Proteus mirabilis, Acinetobacter baumannii, and Haemophilus influenzae. Experimentally verified T11SS-dependent cargo include transferrin-binding protein B (TbpB), the hemophilin homologs heme receptor protein C (HrpC), hemophilin A (HphA), the immune evasion protein factor-H binding protein (fHbp), and the host symbiosis factor nematode intestinal localization protein C (NilC). Here, we examined the specificity of T11SS systems for their cognate cargo proteins using taxonomically distributed homolog pairs of T11SS and hemophilin cargo and explored the ligand binding ability of those hemophilin cargo homologs. In vivo expression in Escherichia coli of hemophilin homologs revealed that each is secreted in a specific manner by its cognate T11SS protein. Sequence analysis and structural modeling suggest that all hemophilin homologs share an N-terminal ligand-binding domain with the same topology as the ligand-binding domains of the Haemophilus haemolyticus heme binding protein (Hpl) and HphA. We term this signature feature of this group of proteins the hemophilin ligand-binding domain. Network analysis of hemophilin homologs revealed five subclusters and representatives from four of these showed variable heme-binding activities, which, combined with sequence-structure variation, suggests that hemophilins are diversifying in function.

59 BASIC BIOLOGICAL SCIENCES↗
NASA NTRS1997

Strain activation of bovine aortic smooth muscle cell proliferation and alignment: study of strain dependency and the role of protein kinase A and C signaling pathways

Smooth muscle cell (SMC) phenotype can be altered by physical forces as demonstrated by cyclic strain-induced changes in proliferation, orientation, and secretion of macromolecules. However, the magnitude of strain required and the intracellular coupling pathways remain ill defined. To examine the strain requirements for SMC proliferation, we selectively seeded bovine aortic SMC either on the center or periphery of silastic membranes which were deformed with 150 mm Hg vacuum (0-7% center; 7-24% periphery). SMC located in either the center or peripheral regions showed enhanced proliferation compared to cells grown under the absence of cyclic strain. Moreover, SMC located in the center region demonstrated significantly (P < 0.005) greater proliferation as compared to those in the periphery. In contrast, SMC exposed to high strain (7-24%) demonstrated alignment perpendicular to the strain gradient, whereas SMC in the center (0-7%) remained aligned randomly. To determine the mechanisms of these phenomena, we examined the effect of cyclic strain on bovine aortic SMC signaling pathways. We observed strain-induced stimulation of the cyclic AMP pathway including adenylate cyclase activity and cyclic AMP accumulation. In addition, exposure of SMC to cyclic strain caused a significant increase in protein kinase C (PKC) activity and enzyme translocation from the cytosol to a particulate fraction. Further study was conducted to examine the effect of strain magnitude on signaling, particularly protein kinase A (PKA) activity as well as cAMP response element (CRE) binding protein levels. We observed significantly (P < 0.05) greater PKA activity and CRE binding protein levels in SMC located in the center as compared to the peripheral region. However, inhibition of PKA (with 10 microM Rp-cAMP) or PKC (with 5-20 ng/ml staurosporine) failed to alter either the strain-induced increase in SMC proliferation or alignment. These data characterize the strain determinants for activation of SMC proliferation and alignment. Although strain activated both the AC/cAMP/PKA and the PKC pathways in SMC, singular inhibition of PKA and PKC failed to prevent strain-induced alignment and proliferation, suggesting either their lack of involvement or the multifactorial nature of these responses.

Non-NASA Center↗
DOE OSTI2022

AI-Based Protein Interaction Screening and Identification (AISID)

In this study, we presented an AISID method extending AlphaFold-Multimer’s success in structure prediction towards identifying specific protein interactions with an optimized AISIDscore. The method was tested to identify the binding proteins in 18 human TNFSF (Tumor Necrosis Factor superfamily) members for each of 27 human TNFRSF (TNF receptor superfamily) members. For each TNFRSF member, we ranked the AISIDscore among the 18 TNFSF members. The correct pairing resulted in the highest AISIDscore for 13 out of 24 TNFRSF members which have known interactions with TNFSF members. Out of the 33 correct pairing between TNFSF and TNFRSF members, 28 pairs could be found in the top five (including 25 pairs in the top three) seats in the AISIDscore ranking. Surprisingly, the specific interactions between TNFSF10 (TNF-related apoptosis-inducing ligand, TRAIL) and its decoy receptors DcR1 and DcR2 gave the highest AISIDscore in the list, while the structures of DcR1 and DcR2 are unknown. The data strongly suggests that AlphaFold-Multimer might be a useful computational screening tool to find novel specific protein bindings. This AISID method may have broad applications in protein biochemistry, extending the application of AlphaFold far beyond structure predictions.

59 BASIC BIOLOGICAL SCIENCES↗