CNVgears: CNVgears: A package to analyze CNVs calling/segmentation...

In SinomeM/CNVgears: A Framework to Combine, Analize and Interpret CNVs Calling Results

Description

CNvgears provides several functions to analyze the results of CNVs calling and/or segmentation on SNPs arrays or NGS data.

Details

The CNVgears package provides several functions useful in order to perform a series of analysis the result of CNVs calling or segmentation pipelines or algorithms, on both Ilummina SNP array (e.g. PennCNV, iPattern or EnsembleCNV) and NGS data (e.g. ModSeg and gCNV pipelines from GATK), in an integrated framework. To do so all the data is imported in a standardized manner, allowing the user to perform analysis and data manipulation regardless of the initial raw data type, from (among the others) CNVRs creation and exclusion of immunoglobulin regions, to de novo CNVs discovery and genic content annotation.

It has been originally developed for the CNVs characterization of the Italian Autism Network (ITAN) collection (DOI: 10.1186/s12888-018-1937-y).

Analysis pipelines examples

Here are briefly illustrated some workflow examples that can be done either interactively on sequentially. See the vignettes for further details.

Staring from the results of gCNV and ModSeg pipelines on WES data in a cohort of families:

1. load the intervals list (using read_NGS_intervals);

2. load samples table with minimal metadata (sample ID, sex, role, family ID);

3. load the segmentation results of all the samples in the cohort, for each pipeline separately;

4. merge adjacent segments (with equal CN);

5. filter out CNVs in immunoglobulin (IG) regions;

6. find eventual oversegmented samples (can be marked or excluded from the analysis);

7. find replicated segments in the pipelines and merge the results into a single data.table;

8. create the CNVRs;

9. exclude common CNVs based on the CNVRs frequency;

10. annotate genic contents of the CNVs

11. find the inheritance pattern of a selected subset of events (or the whole dataset) in the offspring, based on the segments of the parents;

12. fine-screen putative de novo calls using the per-interval raw data (copy ratio or LRR like) of the trio;

13. visualize the good de novo candidate per point raw data in the family to visually confirm the inheritance pattern.

CNVgears functions

The CNVgears functions are organized in groups:

• input/output

• filtering

• CNVRs

• inter results comparison and merging

• de novo discovery/inheritance pattern detection

• plotting

SinomeM/CNVgears documentation built on Nov. 21, 2021, 5:34 a.m.