Machine-learned analysis of the association of next-generation sequencing–based human TRPV1 and TRPA1 genotypes with the sensitivity to heat stimuli and topically applied capsaicin

Abstract

Supplemental Digital Content is Available in the Text.Machine-learning methods associated next generation–based genotypes with the response to hypersensitization to heat induced by topical application of capsaicin.

Machine-learned analysis of the association of next-generation sequencing–based genotypes with persistent pain after breast cancer surgery

“…The analytical protocol for NGS-based genotype vs phenotype association has recently been established in similarly sized cohorts carrying or not carrying a pain-related phenotype. 63 An overview of the data analysis is shown in ½F1 Fig. 1.…”

“…117 Two different commercial NGS platforms are currently widely used for diagnostic purposes: the MiSeq/HiSeq/ NextSeq (Illumina, Hayward, CA) and the IonTorrent PGM (Life Technologies, Carlsbad, CA). Although both platforms differ in the technical implementation of NGS, 63,64 multiple studies have shown that both NGS platforms provide reliable sequencing results in routine clinical diagnostics. 22,94,128…”

“…41,119 This was performed in 5 steps as described previously in all detail. 63 In brief, first, variants found in the whole patient cohort at allelic frequencies below 10% were omitted. Second, variants for which the distribution of homozygous and heterozygous carriers differed from expectation, according to the Hardy-Weinberg equilibrium, 45 were omitted.…”

“…The analysis followed the same concept as applied in a previous analogous assessment previously described in detail. 63 In brief, the relationship between the genotypic and the phenotypic data structure was first explored using unsupervised machine learning to identify a genetic marker pattern that is compatible with the class structure of the phenotypes. Following establishment that the genetic data contained a structure onto which the phenotype class structure can be superimposed to a statistically significant degree, mapping of the genetic parameters to the phenotype classes was performed.…”

“…DNA samples and pain data were available from a cohort of 1000 women 58 among whom 70 had been diagnosed with persistent pain, based on ratings acquired up to 36 months after surgery. 85 A recent data-driven analytical approach was used 63 in which mainly machine-learning algorithms for classifier building were used to uncover the genetic background of persistent pain after breast cancer surgery, although without explicitly aiming to identify a genetic biomarker for clinical use. Techniques of feature selection 41,119 were used to rank the genes in terms of the relative importance of genetic variants, thereby training machine-learned algorithms to perform genotype vs phenotype associations.…”