Qianqian Zhu

A Genome-wide Comparison of the Functional Properties of Rare and Common Genetic Variants in Humans

One of the longest running debates in evolutionary biology concerns the kind of genetic variation that is primarily responsible for phenotypic variation in species. Here, we address this question for humans specifically from the perspective of population allele frequency of variants across the complete genome, including both coding and noncoding regions. We establish simple criteria to assess the likelihood that variants are functional based on their genomic locations and then use whole-genome sequence data from 29 subjects of European origin to assess the relationship between the functional properties of variants and their population allele frequencies. We find that for all criteria used to assess the likelihood that a variant is functional, the rarer variants are significantly more likely to be functional than the more common variants. Strikingly, these patterns disappear when we focus on only those variants in which the major alleles are derived. These analyses indicate that the majority of the genetic variation in terms of phenotypic consequence may result from a mutation-selection balance, as opposed to balancing selection, and have direct relevance to the study of human disease.

Using ERDS to infer copy-number variants in high-coverage genomes.

Although there are many methods available for inferring copy-number variants (CNVs) from next-generation sequence data, there remains a need for a system that is computationally efficient but that retains good sensitivity and specificity across all types of CNVs. Here, we introduce a new method, estimation by read depth with single-nucleotide variants (ERDS), and use various approaches to compare its performance to other methods. We found that for common CNVs and high-coverage genomes, ERDS performs as well as the best method currently available (Genome STRiP), whereas for rare CNVs and high-coverage genomes, ERDS performs better than any available method. Importantly, ERDS accommodates both unique and highly amplified regions of the genome and does so without requiring separate alignments for calling CNVs and other variants. These comparisons show that for genomes sequenced at high coverage, ERDS provides a computationally convenient method that calls CNVs as well as or better than any currently available method.

Genome-wide association study of interferon-related cytopenia in chronic hepatitis C patients

BACKGROUND & AIMS Interferon-alfa (IFN)-related cytopenias are common and may be dose-limiting. We performed a genome wide association study on a well-characterized genotype 1 HCV cohort to identify genetic determinants of peginterferon-α (pegIFN)-related thrombocytopenia, neutropenia, and leukopenia. METHODS 1604/3070 patients in the IDEAL study consented to genetic testing. Trial inclusion criteria included a platelet (Pl) count ≥80×10(9)/L and an absolute neutrophil count (ANC) ≥1500/mm(3). Samples were genotyped using the Illumina Human610-quad BeadChip. The primary analyses focused on the genetic determinants of quantitative change in cell counts (Pl, ANC, lymphocytes, monocytes, eosinophils, and basophils) at week 4 in patients >80% adherent to therapy (n=1294). RESULTS 6 SNPs on chromosome 20 were positively associated with Pl reduction (top SNP rs965469, p=10(-10)). These tag SNPs are in high linkage disequilibrium with 2 functional variants in the ITPA gene, rs1127354 and rs7270101, that cause ITPase deficiency and protect against ribavirin (RBV)-induced hemolytic anemia (HA). rs1127354 and rs7270101 showed strong independent associations with Pl reduction (p=10(-12), p=10(-7)) and entirely explained the genome-wide significant associations. We believe this is an example of an indirect genetic association due to a reactive thrombocytosis to RBV-induced anemia: Hb decline was inversely correlated with Pl reduction (r=-0.28, p=10(-17)) and Hb change largely attenuated the association between the ITPA variants and Pl reduction in regression models. No common genetic variants were associated with pegIFN-induced neutropenia or leucopenia. CONCLUSIONS Two ITPA variants were associated with thrombocytopenia; this was largely explained by a thrombocytotic response to RBV-induced HA attenuating IFN-related thrombocytopenia. No genetic determinants of pegIFN-induced neutropenia were identified.

Interleukin 28B polymorphisms are the only common genetic variants associated with low-density lipoprotein cholesterol (LDL-C) in genotype-1 chronic hepatitis C and determine the association between LDL-C and treatment response

Summary.  Low‐density lipoprotein cholesterol (LDL‐C) levels and interleukin 28B (IL28B) polymorphism are associated with sustained viral response (SVR) to peginterferon/ribavirin (pegIFN/RBV) for chronic hepatitis C (CHC) infection. IL28B has been linked with LDL‐C levels using a candidate gene approach, but it is not known whether other genetic variants are associated with LDL‐C, nor how these factors definitively affect SVR. We assessed genetic predictors of serum lipid and triglyceride levels in 1604 patients with genotype 1 (G1) chronic hepatitis C virus (HCV) infection by genome‐wide association study and developed multivariable predictive models of SVR. IL28B polymorphisms were the only common genetic variants associated with pretreatment LDL‐C level in Caucasians (rs12980275, P = 4.7 × 10−17, poor response IL28B variants associated with lower LDL‐C). The association was dependent on HCV infection, IL28B genotype was no longer associated with LDL‐C in SVR patients after treatment, while the association remained significant in non‐SVR patients (P < 0.001). LDL‐C was significantly associated with SVR for heterozygous IL28B genotype patients (P < 0.001) but not for homozygous genotypes. SVR modelling suggested that IL28B heterozygotes with LDL‐C > 130 mg/dL and HCV RNA ≤600 000 IU/mL may anticipate cure rates >80%, while the absence of these two criteria was associated with an SVR rate of <35%. IL28B polymorphisms are the only common genetic variants associated with pretreatment LDL‐C in G1‐HCV. LDL‐C remains significantly associated with SVR for heterozygous IL28B genotype patients, where LDL‐C and HCV RNA burden may identify those patients with high or low likelihood of cure with pegIFN/RBV therapy.

Systematic assessment of imputation performance using the 1000 Genomes reference panels

Genotype imputation has been widely adopted in the postgenome-wide association studies (GWAS) era. Owing to its ability to accurately predict the genotypes of untyped variants, imputation greatly boosts variant density, allowing fine-mapping studies of GWAS loci and large-scale meta-analysis across different genotyping arrays. By leveraging genotype data from 90 whole-genome deeply sequenced individuals as the evaluation benchmark and the 1000 Genomes Project data as reference panels, we systematically examined four important issues related to genotype imputation practice. First, in a study of imputation accuracy, we found that IMPUTE2 and minimac have the best imputation performance among the three popular imputing software evaluated and that using a multi-population reference panel is beneficial. Second, the optimal imputation quality cutoff for removing poorly imputed variants varies according to the software used. Third, the major contributing factors to consistently poor imputation are low variant heterozygosity, high sequence similarity to other genomic regions, high GC content, segmental duplication and being far from genotyping markers. Lastly, in an evaluation of the imputability of all known GWAS regions, we found that GWAS loci associated with hematological measurements and immune system diseases are harder to impute, as compared with other human traits. Recommendations made based on the above findings may provide practical guidance for imputation exercise in future genetic studies.