CNVds

Description

CNVds is an R package developed to evaluate the pathogenicities of human autosomal copy number variations (CNVs) based on the dosage sensitivity scores of the genes that are encompassed within. Dosage sensitivity (DS) refers to the phenotypic changes a gene may produce due to changes in its copy numbers. This package accepts inputs in the format of CNV calls, which includes the chromosome number, start & end position, type of variation (deletion ‘DEL’ or duplication ‘DUP’), and the number of copies. Such information can be obtains using existing CNV detection tools like PennCNV. In addition, this package can also accept pre-defined lists of genes by gene symbols (e.g. ‘BRCA1’, ‘CDC45’) for matching genes to their corresponding DS scores.

This package focuses on the following three score metrics:

• Probability of loss intolerance (pLI): the probability that a gene is intolerant to a loss of function mutation.
• Probability of haploinsufficiency (pHI): the probability that a gene is sensitive to copy number loss.
• Probability of triplosensitivity (pTS): the probability that a gene is sensitive to copy number gain.

The reference DS scores are based on recent publications of genome-wide dosage sensitivity mappings. Compared to solely measuring the physical attributes of the CNVs (e.g. size, number of genes, or known association with disease), DS scores provide additional metrics that are useful in the systematic identification of pathogenic CNVs. The CNVds package was developed using R version 4.2.1 (2022-06-23 ucrt), Platform: x86_64-windows-10_x64, and running under: version build 19044.

Installation

You can install the development version of CNVds from GitHub with:

require("devtools")
devtools::install_github("jenydu/CNVds", build_vignettes = TRUE)
library("CNVds")

To run the Shiny app:

runCNVds()

Overview

ls("package:CNVds")
data(package = "CNVds") 
browseVignettes("CNVds")

CNVds currently contains 8 functions:

• annotateCNV(): Given a CNV call, return all genes that are fully contained in this region based on the GRCh37 or GRCh38 reference genome (specified by user).
• findpLI(): Given a list of genes, find the corresponding pLI (prob. of loss intolerance) scores for each gene.
• findpHI(): Given a list of genes, find the corresponding pHI (prob. of haploinsufficiency) scores for each gene.
• findpTS(): Given a list of genes, find the corresponding pTS (prob. of triplosensitivity) scores for each gene.
• geneDSscores(): Given a single gene symbol, find its associated pLI, pHI, and pTS scores.
• geneNoScores(): Given a list of genes and the name of the DS score (pLI/pHI/pTS), return the list of genes that don’t have a score in the corresponding reference table, and print out the percentage of genes without a score. (Note: it is recommended that you run this function before finding the dosage sensitivity scores, for sanity check purposes.)
• plotCNVcounts(): Given a list of CNV calls, plot the total number of unique duplication and deletion CNV regions in each chromosome.
• plotScoresByChr(): Given a list of genes and their corresponding DS scores (pLI/pHI/pTS), plot the distribution of the scores and label genes that are especially dosage-sensitive (i.e. above a certain user-defined threshold).

Contributions

The author of the package is Jun Ni Du. The approach to dosage sensitivity score calculation is based on Alexander-Bloch, A. et al. (2022)’s publication, with no direct codes taken. The reference tables are obtained from the following publications and/or packages:

• grch37.rda and grch38.rda are obtained from the annotables package. The source files are published by Ensembl (Cunningham, F. et al. (2022)). They are subsets of the original datasets to only include the gene symbol, chromosome number, and the start & end position of all autosomal genes.
• pHaplo_pTriplo_data.rda is obtained from Collins, R. L. et al. (2022).
• pLI_data.rda is obtained from Karczewski, K. J. et al. (2020).

A sample input of 200 CNV calls, sampleInputCNV.csv, is included for package demonstration purposes. It is randomly subsetted from Ming, C. et al. (2021).

The plotScoresByChr and plotCNVcounts functions uses the ggplot function from the ggplot2 R package. The plotScoresByChr funciton also uses the drop_na function from the tidyr R package, and the setNames function from the stats R package. The runCNVds function uses the shiny and shinyalert R packages to build the shiny app. Its implementation adapted codes from the “009-upload” Shiny example, the runTestingPackage function from the TestingPackage R package, and code snippets from online forums such as Stack Overflow (see References section for a complete list of sources; additional details can also be found in inst/shiny-scripts/app.R).

Other than the above mentioned, the rest of functions are implemented solely by the author.

References

Alexander-Bloch, A. et al. (2022). Copy number variant risk scores associated with cognition, psychopathology, and brain structure in youths in the philadelphia neurodevelopmental cohort. JAMA Psychiatry 79, 699–709. https://doi.org/10.1001/jamapsychiatry.2022.1017

Collins, R. L. et al. (2022). A cross-disorder dosage sensitivity map of the human genome. Cell 185, 3041-3055.e25. https://doi.org/10.1016/j.cell.2022.06.036

Cunningham, F. et al. (2022). Ensembl 2022. Nucleic Acids Research 50(1), D988-D995. https://doi.org/10.1093/nar/gkab1049

Jgraph (2022). Drawio-desktop. Image created by Du, J. Retrieved November 14, 2022, from https://app.diagrams.net/

Karczewski, K. J. et al. (2020). The mutational constraint spectrum quantified from variation in 141,456 humans. Nature 581, 434–443. https://doi.org/10.1038/s41586-020-2308-7

Ming, C. et al. (2021). Whole genome sequencing–based copy number variations reveal novel pathways and targets in Alzheimer’s disease. Alzheimer’s & Dementia 18, 1846-1867. https://doi.org/10.1002/alz.12507

R Core Team (2020). R: A language and environment for statistical computing. R Foundation for Statistical Computing, Vienna, Austria. ISBN 3-900051-07-0. http://www.R-project.org/

RStudio (2018). rstudio/shiny-examples: 009-upload. Github. https://github.com/rstudio/shiny-examples/tree/main/009-upload

Silva, A. (2022). Anjalisilva/TestingPackage: A Simple R Package Illustrating Components of an R Package: 2019-2022 BCB410H - Applied Bioinformatics, University of Toronto, Canada. GitHub. https://github.com/anjalisilva/TestingPackage

Thothal. (2019). Use default csv file when no file uploaded to shiny app. Stack Overflow. https://stackoverflow.com/questions/55566874/use-default-csv-file-when-no-file-uploaded-to-shiny-app

Turner, S. (2022). Stephenturner/annotables: R data package for annotating/converting Gene IDs. Github. https://github.com/stephenturner/annotables

Wickham, H. (2016). ggplot2: Elegant Graphics for Data Analysis. Springer-Verlag New York. ISBN 978-3-319-24277-4. https://ggplot2.tidyverse.org

Wickham, H., Girlich, M. (2022). tidyr: Tidy Messy Data. https://tidyr.tidyverse.org

Acknowledgements

This package was developed as part of an assessment for 2022 BCB410H: Applied Bioinformatics course at the University of Toronto, Toronto, CANADA. CNVds welcomes issues, enhancement requests, and other contributions. To submit an issue, use the GitHub issues. Many thanks to those who provided feedback to improve this package.