A weighted burden test using logistic regression for integrated analysis of sequence variants, copy number variants and polygenic risk score. (January 2019)
- Record Type:
- Journal Article
- Title:
- A weighted burden test using logistic regression for integrated analysis of sequence variants, copy number variants and polygenic risk score. (January 2019)
- Main Title:
- A weighted burden test using logistic regression for integrated analysis of sequence variants, copy number variants and polygenic risk score
- Authors:
- Curtis, David
- Abstract:
- Abstract Previously described methods of analysis allow variants in a gene to be weighted more highly according to rarity and/or predicted function and then for the variant contributions to be summed into a gene-wise risk score, which can be compared between cases and controls using at -test. However, this does not allow incorporating covariates into the analysis. Schizophrenia is an example of an illness where there is evidence that different kinds of genetic variation can contribute to risk, including common variants contributing to a polygenic risk score (PRS), very rare copy number variants (CNVs) and sequence variants. A logistic regression approach has been implemented to compare the gene-wise risk scores between cases and controls, while incorporating as covariates population principal components, the PRS and the presence of pathogenic CNVs and sequence variants. A likelihood ratio test is performed, comparing the likelihoods of logistic regression models with and without this score. The method was applied to an ethnically heterogeneous exome-sequenced sample of 6000 controls and 5000 schizophrenia cases. In the raw analysis, the test statistic is inflated but inclusion of principal components satisfactorily controls for this. In this dataset, the inclusion of the PRS and effect from CNVs and sequence variants had only small effects. The set of genes which areFMRP targets showed some evidence for enrichment of rare, functional variants among cases (p = 0.0005). ThisAbstract Previously described methods of analysis allow variants in a gene to be weighted more highly according to rarity and/or predicted function and then for the variant contributions to be summed into a gene-wise risk score, which can be compared between cases and controls using at -test. However, this does not allow incorporating covariates into the analysis. Schizophrenia is an example of an illness where there is evidence that different kinds of genetic variation can contribute to risk, including common variants contributing to a polygenic risk score (PRS), very rare copy number variants (CNVs) and sequence variants. A logistic regression approach has been implemented to compare the gene-wise risk scores between cases and controls, while incorporating as covariates population principal components, the PRS and the presence of pathogenic CNVs and sequence variants. A likelihood ratio test is performed, comparing the likelihoods of logistic regression models with and without this score. The method was applied to an ethnically heterogeneous exome-sequenced sample of 6000 controls and 5000 schizophrenia cases. In the raw analysis, the test statistic is inflated but inclusion of principal components satisfactorily controls for this. In this dataset, the inclusion of the PRS and effect from CNVs and sequence variants had only small effects. The set of genes which areFMRP targets showed some evidence for enrichment of rare, functional variants among cases (p = 0.0005). This approach can be applied to any disease in which different kinds of genetic and non-genetic risk factors make contributions to risk. … (more)
- Is Part Of:
- European journal of human genetics. Volume 27:Number 1(2019)
- Journal:
- European journal of human genetics
- Issue:
- Volume 27:Number 1(2019)
- Issue Display:
- Volume 27, Issue 1 (2019)
- Year:
- 2019
- Volume:
- 27
- Issue:
- 1
- Issue Sort Value:
- 2019-0027-0001-0000
- Page Start:
- 114
- Page End:
- 124
- Publication Date:
- 2019-01
- Subjects:
- Human genetics -- Periodicals
Medical genetics -- Periodicals
616.042 - Journal URLs:
- http://www.nature.com/ejhg/index.html ↗
https://www.karger.com/Journal/Home/224162 ↗
http://www.nature.com/ ↗ - DOI:
- 10.1038/s41431-018-0272-6 ↗
- Languages:
- English
- ISSNs:
- 1018-4813
- Deposit Type:
- Legaldeposit
- View Content:
- Available online (eLD content is only available in our Reading Rooms) ↗
- Physical Locations:
- British Library DSC - 3829.730020
British Library DSC - BLDSS-3PM
British Library HMNTS - ELD Digital store - Ingest File:
- 12703.xml