How to choose good tools for identifying functional SNPs?

Question

[arr234]

If you look at the materials and methods of (https://www.ncbi.nlm.nih.gov/pmc/articles/PMC5449402/) this paper, they have suggested using any three tools from each method (sequence homology-based methods, supervised learning methods, protein sequence and structure-based methods and consensus-based methods) to find the deleterious SNP. How do I choose the three best tools for each method?

Answer 1

[Kevin Blighe]

Take your pick...

Sequence and evolutionary- / homology- / conservation-based methods (for missense predictions)

• SIFT

• Align-GVGD

• Mutation Assessor

• PANTHER
• MAPP
• MUpro
• MutationTasster
• I-Mutant
• KD4v
• FATHMM
• dbNSFP

Protein sequence and structure-based methods (for missense predictions)

• PolyPhen-2
• LS-SNP/PDB
• SNPeffect 4.0

Supervised-learning methods (for missense predictions)

• PMut
• SNAP
• PhD-SNP
• SNPs&GO
• Parepro
• CanPredict
• nsSNPAnalyzer
• MutPred2
• HANSA
• MutationTaster

Splice predictions

• NNSplice
• GeneSplicer
• MaxEntScan
• HSF
• NetGene2

Protein modelling (from amino acid sequence)

• Protein Model Portal

[uses various modelling algorithms and produces PDB files, which can be loaded into protein viewers like Jmol]

Non-coding (i.e. regulatory)

• CADD (germline variants)
• DANN (germline variants)
• FATHMM-MKL (germline variants)
• GWAVA (germline variants | somatic mutations)
• Funseq2 (somatic mutations)
• SurfR (rare variants | complex disease variants | all other variants)

Other

• GWAS3D / GWAS4D

---

For further reading:

• UK perspective, refer to the guidelines by the Association for Clinical Genomic Science: http://www.acgs.uk.com/quality/best-practice-guidelines/
• US perspective, refer to American College of Medical Genetics: https://www.acmg.net/docs/Standards_Guidelines_for_the_Interpretation_of_Sequence_Variants.pdf

------------------------------------------------------------

I assure you that there are even more tools than these. These are categorised to the best of my and my colleague's knowledge.

Previous posting: A: pathogenicity predictors of cancer mutations

Kevin

[jared.andrews07]

Great list. Don't forget the Variant Effect Predictor (Coding/Protein, mostly), SnpEff (variety of fields), and DeepSea (Non-coding/Regulatory).

[arr234]

Thank you, Kevin Blighe. So now I can choose randomly any tool from each category in this list or should the choice be based on any criteria for best tools?

[Kevin Blighe]

You could take a look through each and try to make your choices that way.

[Kevin Blighe]

Take a look at the material under 'further reading'. For example, in the UK, the clinical genetics regulatory body recommends certain tools over others.

[Mehulsharma.253]

what tools would you recommend to asses the effect of variants on ligand binding ?Schrodinger maestro seems good but it's a paid software.

[Kevin Blighe]

Sounds more like quantum chemistry ... ? Take a look at Spartan molecular modelling, although, you will require a powerful computer to accurately predict the effects. Could also look at the link that I posted under Protein modelling (from amino acid sequence)