What does this research mean for the field?

The optimal combination of in silico prediction tools for classifying long QT syndrome gene variants is gene-dependent, providing high predictive value for KCNQ1 and KCNH2 variants but requiring caution for SCN5A variants. Novelty: ClaimNovelty.NOVEL_FINDING. Consensus alignment: ConsensusAlignment.NEUTRAL.

May 12, 2015BMC Medical GeneticsOpen Access

Assessment of the predictive accuracy of five in silico prediction tools, alone or in combination, and two metaservers to classify long QT syndrome gene mutations

Q: What is the clinical evidence from this study?

Study design: Other. Population: Long QT syndrome gene mutations (n=312). Intervention: In silico prediction tools (SIFT, PROVEAN, SNAP) vs. Functional characterisation (ground truth). Primary outcome: Predictive accuracy for all three LQT genes combined (MCC 0.44).

View Full Paper

Ask AI

Bookmark

Key result

Combining SIFT, PROVEAN, and SNAP classifies long QT syndrome mutations with ~83% accuracy.

MCC 0.44
n=312

Why the study?

Do in silico prediction tools accurately classify long QT syndrome gene mutations compared to functional characterization?

Population

312 missense single nucleotide variants in KCNQ1, KCNH2, and SCN5A genes previously characterized by in…

Comparison

Five in silico prediction tools alone or in… vs Reference standard of functional…

Design

Preclinical

Authors

ILIvone LeongGenomics England ASAlexander StuckeyID Genomics (United States)DLDaniel LaiProvincial Health Services Authority

Discussion

Member takes

Implication

In silico tools may aid LQTS variant classification; leaves open gene-specific validation before clinical adoption.

Structured PICO

Do in silico prediction tools accurately classify long QT syndrome gene mutations compared to functional characterization?

Population

312 missense single nucleotide variants (SNVs) in KCNQ1, KCNH2, and SCN5A genes (283 pathogenic, 29 benign) previously characterized by in vitro or co-segregation studies.

Intervention

Five in silico prediction tools (SIFT, PolyPhen-2, PROVEAN, SNPs&GO, SNAP) alone or in combination, and two metaservers (Meta-SNP, PredictSNP).

Comparator

Reference standard of functional characterization by in vitro or co-segregation studies.

Outcome

Accuracy, sensitivity, specificity, and Matthews Correlation Coefficient (MCC) for predicting pathogenicity.surrogate

Main Result

Effect estimate: MCC 0.44

In silico prediction tools can aid in assessing the pathogenicity of KCNQ1 and KCNH2 variants in Long QT syndrome, but their accuracy is gene-dependent and poor for SCN5A variants.

Limitations

Low number of benign SNVs for all three LQTS genes, which may affect the reliability of benign predictions.
Only a specific subset of in silico prediction tools (SIFT, PolyPhen-2, PROVEAN, SNPs&GO, SNAP, Meta-SNP, and PredictSNP) were investigated.

Cite This Study

Leong et al. (2015) studied Long QT syndrome gene mutations (n=312). In silico prediction tools (SIFT, PROVEAN, SNAP) vs. Functional characterisation (ground truth) was evaluated on Predictive accuracy for all three LQT genes combined (MCC 0.44). The combination of SIFT, PROVEAN, and SNAP provided the best predictive performance for classifying long QT syndrome gene mutations, achieving 82.7% accuracy and an MCC of 0.44.

synapsesocial.com/papers/6a0cd6abeff1ba4410c4015b https://doi.org/10.1186/s12881-015-0176-z

View Full Paper

Ask AI

Bookmark

Assessment of the predictive accuracy of five in silico prediction tools, alone or in combination, and two metaservers to classify long QT syndrome gene mutations

Discussion

Member takes

Implication

Structured PICO

Main Result

Limitations

Cite This Study

Assessment of the predictive accuracy of five in silico prediction tools, alone or in combination, and two metaservers to classify long QT syndrome gene mutations

Discussion

Member takes

Implication

Structured PICO

Main Result

Limitations

Cite This Study

Also Consider

Also Consider