Page 102 - TD-3-4

Tumor Discovery Bioinformatics insights into CCL2 mutations

ensembl.org), which provides comprehensive genomic proteins’ free energy indicates a change in stability, where
information (Ensembl ID: ENSG00000108691). This negative ΔΔG indicates destabilization, whereas positive
32
dataset was chosen due to its extensive collection of ΔΔG indicates stabilization. 27,42 Various computational
genomic variations. In addition, data were obtained from tools, including mCSM, SDM, and DUET, were employed
the UniProtKB database (https://www.uniprot.org), which to assess the impact of mutations on CCL2 protein stability
includes sequence details in FASTA format (UniProt and function, providing valuable insights into how specific
ID: P13500), and the National Center for Biotechnology mutations alter protein behavior. 27
Information (NCBI) (https://www.ncbi.nlm.nih.gov). 33,34 The SDM online tool, available at http://marid.bioc.
Structural information regarding CCL2 was retrieved from cam.ac.uk/sdm2, accurately predicts the impact of
the Protein Data Bank (PDB) (CCL2 PDB ID: 2LIE, and mutations on protein stability through in silico analysis. It
CCR2 PDB ID: 2LWL) through the website http://www. examines variations in amino acid substitutions tolerated
rcsb.org/. 35 by homologous proteins with known 3D structures in

2.2. Computational prediction of mutation specific environments, providing valuable insights into how
pathogenicity mutations can affect protein behavior and stability. 27,43,44
To assess the effects of mutations on CCL2 proteins, we The mCSM computational technique uses machine
used computational methods essential for distinguishing learning to predict the effects of missense mutations on
between single-nucleotide variations that are detrimental, CCL2 protein stability and function. It employs graph-based
benign, or cancerous. Accurate evaluation of the functional signatures to depict how mutations impact the interaction
impacts of these genetic alterations is essential for network among residues. By integrating evolutionary
identifying disease-causing mutations. Meta-SNP (http:// information, complex network metrics, and energetic
snps.biofold.org/meta-snp/) is a web-based tool that uses factors, mCSM creates an accurate predictive model. It is
21,26
multiple predictive algorithms to assess the potential available at http://structure.bioc.cam.ac.uk/mcsm. The
pathogenicity of protein mutations. We employed meta- mCSM method evaluates how mutations affect protein
22
SNP to assess mutation pathogenicity by integrating the structure and interactions with other proteins and nucleic
predictions from algorithms such as SNAP, PhD-SNP, acids. It provides insights into how mutations contribute
36
Panther, and SIFT. 38 to disease development, improving our understanding
37
of the molecular mechanisms of various diseases. This
Meta-SNP is a reliable tool for predicting how mutations computational approach clarifies the effects of mutations
affect protein functions. It combines results from various on protein stability, function, and interactions, enhancing
algorithms using a random forest technique, which involves our understanding of disease pathogenesis. 45,46
100 decision trees within the WEKA library. This method DUET (http://structure.bioc.cam.ac.uk/duet) is a
ensures the precise assessment of a mutation’s impact on its
protein function. The meta-SNP algorithm was trained computational tool that predicts changes in protein stability
39
on the SV-2009 dataset through a 20-fold cross-validation due to individual mutations. It combines two methods to
procedure and provides dependable predictions of disease- calculate the change in folding free energy (ΔΔG in Kcal/
mol), with negative ΔΔG values indicating destabilizing
causing variants. It outputs a probability score between 0 mutations and positive values indicating stabilizing
22
and 1, with scores above 0.5 indicating a higher likelihood mutations. 28,47 DUET uses support vector machines trained
of the mutation being pathogenic. 40
with sequential minimal optimization to integrate the
Next, we used the Polyphen-2 web server (http:// predictions from mCSM and SDM. This machine learning
genetics.bwh.harvard.edu/pph2/), which is an automated technique leverages the strengths of both methods to provide
tool designed to predict the potential effects of amino a comprehensive assessment of how missense mutations
acid substitutions on the structure and function of human affect protein stability. By combining these predictive
proteins. This site features an input form that enables capabilities, DUET enhances the accuracy and reliability of
41
users to query either a single amino acid substitution or mutation impact predictions on protein stability. 48
a coding, non-synonymous SNP that is annotated in the
dbSNP database. 23 2.4. Evaluating cancer-related mutations using the
FATHMM server
2.3. Computational assessment of protein stability
The FATHMM server, available at http://fathmm.
Mutations can affect a protein’s structure and stability, biocompute.org.uk, was used to assess the cancer relevance
resulting in changes in free energy (ΔΔG in Kcal/mol). The of deleterious mutations in CCL2. This web server predicts
difference between the mutated (ΔGm) and original (ΔGw) the functional consequences of both coding and noncoding

Volume 3 Issue 4 (2024) 4 doi: 10.36922/td.3891