Publications

Global characterization of copy number variants in epilepsy patients from whole genome sequencing

• Article in PLoS Genetics
• Manuscript in bioRxiv.
• GitHub repo with code to reproduce the figures and numbers.
• Data on FigShare includes calls and data necessary to reproduce the analysis.

The CNV calls in the epilepsy patients and controls can download be downloaded from the FigShare repo. The file cnvs-PopSV-Epilepsy-198affected-301controls-5kb.tsv.gz (21Mb) is tab-delimited with the following columns:

• chr, start, end: the position in the reference genome.
• project: either affected to denote an epilepsy patient, or control for control individuals.
• sample: the sample name.
• z: the Z-score used for calling the CNV.
• fc: the fold change compared to the coverage in the reference samples. <1 for deletions and >1 for duplications.
• mean.cov: the average coverage in the reference samples.
• pv/qv: the P-value and Q-value (or FDR threshold) for the call.
• nb.bins.cons: the number of consecutive 5 Kbp bins in the call.

Human copy number variants are enriched in regions of low mappability

• Article in Nucleic Acids Research
• Manuscript in bioRxiv
• GitHub repo with code to reproduce the figures and numbers.
• Data on FigShare includes calls and data necessary to reproduce the analysis.

The calls from the three datasets (640 genomes total) were merged and formatted into single files to facilitate external analysis/usage of our CNV catalog. The files are available from the FigShare repository.

• CNV-PopSV-Twin_CageKid_GoNL-germline.tsv file contains all the germline CNV calls for the three datasets.
• CNV-PopSV-CageKid-somatic.tsv contains the somatic CNV calls from the renal cancer dataset.

Both files are tab-delimited text files with columns:

• sample: the sample name.
• chr, start, end: the position in the reference genome.
• fc: the fold change compared to the coverage in the reference samples. <1 for deletions and >1 for duplications.
• qv: the Q-value, or FDR threshold, for the call.
• cn2.dev: deviation from the 2 copy number state, i.e. the absolute number of copies of difference between the variant and a diploid state.
• cn: the estimated copy-number state.

Columns qv and cn2.dev can be used to further select higher confidence calls. For example, one could select calls with small qv and/or high cn2.dev. Copy number estimation (column cn) is only reliable when the the variant spans several bins, hence it’s normal to have partial estimates for small calls. Furthermore, copy number estimation in low-coverage regions is more challenging and calls in these regions often exhibits large cn (or fc).

The regions with novel CNVs (as defined in the article) are listed in CNV-novel-notTGP.tsv. The count/freq columns represent the frequency of CNVs in our cohort.