To improve the analysis of case-parent trios data, Dr. Wan-Yu Lin and her former master’s student, Ms. Yun-Chieh Liang, at National Taiwan University, have proposed an adaptive combination of P-values method. This study has been published in June in the Scientific Reports.
[Photo: Dr. Wan-Yu Lin]
With the development of next-generation sequencing technology, we can now investigate a lot of rare variants in human genome. It has been suspected that rare variants are responsible substantially for the susceptibility to complex diseases, but they are usually difficult to detect due to the extreme sparsity. Most rare-variant association analyses strengthen signals by aggregating the information of multiple variants in a gene/region. “However,” Dr. Lin pointed out, “this pooling strategy inevitably leads to the inclusion of neutral variants, which usually compromises the power of association tests.”
“Our adaptive combination of P-values method removes variants with larger P-values, which are more likely to be neutral,” said Ms. Liang. “Although this method is originally proposed for rare-variant association analyses, it can be generalized to pathway analysis or metabolomics studies.”
The authors conclude that their method generates valid statistical inference in the presence of population stratification, because the test statistic is developed by conditioning on parental genotypes. With regard to statistical methods for next-generation sequencing data analyses, validity may be hampered by population stratification, whereas power may be affected by the inclusion of neutral variants. This method is recommended for its robustness to these two factors: population stratification and the inclusion of neutral variants.