SAAFEC-SEQ: A Sequence-Based Method for Predicting the Effect of Single Point Mutations on Protein Thermodynamic Stability

Li, Gen; Panday, Shailesh Kumar; Alexov, Emil

doi:10.3390/ijms22020606

Cited by 85 publications

(90 citation statements)

References 51 publications

(64 reference statements)

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…As mentioned above, few available methods can predict the effect of the variants on the protein stability starting from sequence only. We therefore compared ACDC-NN-Seq on three datasets with the following sequence-based methods: DDGun [14], INPS [13], I-Mutant2.0 [16], MUpro [17] and the recent SAAFEC-SEQ [15].…”

Section: Comparison With Other Sequence-based Machine-learning Methodsmentioning

confidence: 99%

“…Since experimental measurement of ∆∆G is a time-consuming and complex task, during the last decades several computational tools have been developed to predict ∆∆G values. Some methods are structure-based, requiring the knowledge of the protein tertiary structure [7][8][9][10][11][12], others are sequence-based, either relying only on protein sequences [13][14][15] or optionally taking advantage of the protein structure when available [16,17]. However, most of these methods violate the antisymmetry property and suffer from high biases in predicting reverse variations [10,[18][19][20][21][22][23].…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

A Deep-Learning Sequence-Based Method to Predict Protein Stability Changes Upon Genetic Variations

Pancotti

Benevenuta

Repetto

et al. 2021

Genes

View full text Add to dashboard Cite

Several studies have linked disruptions of protein stability and its normal functions to disease. Therefore, during the last few decades, many tools have been developed to predict the free energy changes upon protein residue variations. Most of these methods require both sequence and structure information to obtain reliable predictions. However, the lower number of protein structures available with respect to their sequences, due to experimental issues, drastically limits the application of these tools. In addition, current methodologies ignore the antisymmetric property characterizing the thermodynamics of the protein stability: a variation from wild-type to a mutated form of the protein structure (XW → XM) and its reverse process (XM → XW) must have opposite values of the free energy difference (ΔΔGWM = – ΔΔGMW). Here we propose ACDC-NN-Seq, a deep neural network system that exploits the sequence information and is able to incorporate into its architecture the antisymmetry property. To our knowledge, this is the first convolutional neural network} to predict protein stability changes relying solely on the protein sequence. We show that ACDC-NN-Seq compares favorably with the existing sequence-based methods.

show abstract

Section: Comparison With Other Sequence-based Machine-learning Methodsmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

A Deep-Learning Sequence-Based Method to Predict Protein Stability Changes Upon Genetic Variations

Pancotti

Benevenuta

Repetto

et al. 2021

Genes

View full text Add to dashboard Cite

show abstract

“…The extracellular domain of OPRM1 was used to calculate the folding free energy changes due to mutations A6V and N40D. They were calculated by the SAAFEC-SEQ algorithm [ 66 ], along with third party webservers such as mCSM Protein Stability Change Upon Mutation web version [ 67 ], SDM(V2) [ 68 ], DUET [ 57 ], CUPSAT [ 58 ], and I-Mutant 3.0 [ 69 ].…”

Section: Methodsmentioning

confidence: 99%

Opioid Addiction and Opioid Receptor Dimerization: Structural Modeling of the OPRD1 and OPRM1 Heterodimer and Its Signaling Pathways

Hand

Alexov

2021

IJMS

Self Cite

View full text Add to dashboard Cite

Opioid addiction is a complex phenomenon with genetic, social, and other components. Due to such complexity, it is difficult to interpret the outcome of clinical studies, and thus, mutations found in individuals with these addictions are still not indisputably classified as opioid addiction-causing variants. Here, we computationally investigated two such mutations, A6V and N40D, found in the mu opioid receptor gene OPRM1. The mutations are located in the extracellular domain of the corresponding protein, which is important to the hetero-dimerization of OPRM1 with the delta opioid receptor protein (OPRD1). The hetero-dimerization of OPRD1–OPRM1 affects the signaling pathways activated by opioids and natural peptides and, thus, could be considered a factor contributing to addiction. In this study, we built four 3D structures of molecular pathways, including the G-protein signaling pathway and the β-arrestin signaling pathway of the heterodimer of OPRD1–OPRM1. We also analyzed the effect of mutations of A6V and N40D on the stability of individual OPRM1/OPRD1 molecules and the OPRD1–OPRM1 heterodimer with the goal of inferring their plausible linkage with opioid addiction. It was found that both mutations slightly destabilize OPRM1/OPRD1 monomers and weaken their association. Since hetero-dimerization is a key step for signaling processes, it is anticipated that both mutations may be causing increased addiction risk.

show abstract

“…Neutral or benign effects were not considered in the study (Table 1). [36]. Default settings (Temperature: 25 0 C, pH: 7.0) and the "DDG Value and Binary Classification" option were preferred in I-Mutant.…”

Section: Data Extraction and Evaluation Of The Missense Snps Of Ifitm1 Ifitm2 And Ifitm3mentioning

confidence: 99%

Elucidation of the interactions between SARS-CoV-2 Spike protein and wild and mutant types of IFITM proteins by in silico methods

Giritlioğlu

KÜÇÜK>

2021

Preprint

View full text Add to dashboard Cite

COVID-19 is a viral disease that has been a threat to the whole world since 2019. Although effective vaccines against the disease have been developed, there are still points to be clarified about the mechanism of SARS-CoV-2, which is the causative agent of COVID-19. In this study, we determined the binding energies and the bond types of complexes formed by open (6VYB) and closed (6VXX) forms of the Spike protein of SARS-CoV-2 and wild and mutant forms of IFITM1, IFITM2, and IFITM3 proteins using the molecular docking approach. First, all missense SNPs were found in the NCBI Single Nucleotide Polymorphism database (dbSNP) for IFITM1, IFITM2, and IFITM3 and analyzed with SIFT, PROVEAN, PolyPhen-2, SNAP2, Mutation Assessor, and PANTHER cSNP web-based tools to determine their pathogenicity. When at least four of these analysis tools showed that the SNP had a pathogenic effect on the protein product, this SNP was saved for further analysis. Delta delta G (DDG) and protein stability analysis for amino acid changes were performed in the web-based tools I-Mutant, MUpro, and SAAFEC-SEQ. The structural effect of amino acid change on the protein product was made using the HOPE web-based tool. HawkDock server was used for molecular docking and Molecular Mechanics/Generalized Born Surface Area (MM/GBSA) analysis and binding energies of all complexes were calculated. BIOVIA Discovery Studio program was utilized to visualize the complexes. Hydrogen bonds, salt bridges, and non-bonded contacts between Spike and IFITM protein chains in the complexes were detected with the PDBsum web-based tool. The best binding energy among the 6VYB-IFITM wild protein complexes belong to 6VYB-IFITM1 (-46.16 kcal/mol). Likewise, among the 6VXX-IFITM wild protein complexes, the most negative binding energy belongs to 6VXX-IFITM1 (-52.42 kcal/mol). An interesting result found in the study is the presence of hydrogen bonds between the cytoplasmic domain of the IFITM1 wild protein and the S2 domain of 6VYB. Among the Spike-IFITM mutant protein complexes, the best binding energy belongs to the 6VXX-IFITM2 N63S complex (-50.77 kcal/mol) and the worst binding energy belongs to the 6VXX-IFITM3 S50T complex (4.86 kcal/mol). The study suggests that IFITM1 protein may act as a receptor for SARS-CoV-2 Spike protein. Assays must be advanced from in silico to in vitro for the determination of the receptor-ligand interactions between IFITM proteins and SARS-CoV-2.

show abstract

SAAFEC-SEQ: A Sequence-Based Method for Predicting the Effect of Single Point Mutations on Protein Thermodynamic Stability

Cited by 85 publications

References 51 publications

A Deep-Learning Sequence-Based Method to Predict Protein Stability Changes Upon Genetic Variations

A Deep-Learning Sequence-Based Method to Predict Protein Stability Changes Upon Genetic Variations

Opioid Addiction and Opioid Receptor Dimerization: Structural Modeling of the OPRD1 and OPRM1 Heterodimer and Its Signaling Pathways

Elucidation of the interactions between SARS-CoV-2 Spike protein and wild and mutant types of IFITM proteins by in silico methods

Contact Info

Product

Resources

About