Quantitative prediction of the effect of genetic variation using hidden Markov models

Abstract

Background: With the development of sequencing technologies, more and more sequence variants are available for investigation. Different classes of variants in the human genome have been identified, including single nucleotide substitutions, insertion and deletion, and large structural variations such as duplications and deletions. Insertion and deletion (indel) variants comprise a major proportion of human genetic variation. However, little is known about their effects on humans. The absence of understanding is largely due to the lack of both biological data and computational resources.

Results: This paper presents a new indel functional prediction method HMMvar based on HMM profiles, which capture the conservation information in sequences. The results demonstrate that a scoring strategy based on HMM profiles can achieve good performance in identifying deleterious or neutral variants for different data sets, and can predict the protein functional effects of both single and multiple mutations.

Conclusions: This paper proposed a quantitative prediction method, HMMvar, to predict the effect of genetic variation using hidden Markov models. The HMM based pipeline program implementing the method HMMvar is freely available at https://bioinformatics.cs.vt.edu/zhanglab/hmm.

Figure 1

A pipeline of variant prediction using HMMvar.

Figure 2

HMMvar score distribution of the dbSNP dataset. (a) Histogram of HMMvar scores for disease associated indels and nondisease associated indels. (b) Distribution of sample means of HMMvar scores from the two categories (LSDB and nonLSDB).

Figure 3

Distributions of HMMvar scores for different types of variants.

Figure 4

Compare HMMvar prediction with SIFT Indel prediction on dbSNP indel dataset. Distributions of HMMvar of indels that are predicted as damaging (left) and neutral (right) by SIFT Indel.

Figure 5

HMMvar and Provean score distributions and mean/error bars of TP53 mutations binned into 15 classes in terms of transactivity level. (a) HMMvar score distribution of the 15 classes (x-axis represents the 15 classes based on the median of transactivity levels). (b) Provean score distribution of the 15 classes. (c) Mean along with error bar of HMMvar scores in each class. (d) Mean along with error bar of Provean scores in each class.

Figure 6

ROC curve and standard error of the HMMvar score and the Provean score. (a) ROC curve of the Provean score and the HMMvar score to distinguish "nonfunctional" and "partly functional" classes from "functional" and "supertrans" classes. (b) Standard error of the mean of Provean and HMMvar scores in the 15 transactivity level classes.

Figure 7

The HMMvar score of TP53 variants grouped by SIFT SNP prediction.

Figure 8

The relationship between the HMMvar score and the position of an artificially introduced variant.