Release 3.0.0, June 20th, 2023
Major update: the VarCoPP2.0 model for GRCh38/hg38 is added!
From now on, you should choose the genome assembly (GRCh37/hg19 or GRCh38/hg38) to which your variants correspond in the Input form.
A few minor new features and improvements are also included:
- For variants in GRCh38/hg38, the cDNA change information is provided in the "Annotations" part of the "Digenic combination details" page (given that it was available in CADD v1.6);
- We improved the way InDels are parsed from the variant input, so that all exceptional cases can be handled;
- The Job summary pane was augmented with a breakdown of the variants discarded in the pre-processing.
You can see how many variants either had an invalid zygosity or were filtered away because of input filter settings or missing CADD score.
It is possible to download a .txt file with the variant keys for each category.
If you can share your variant data, don't hesitate to send us the variants with missing CADD score so that we can try to add them to our database;
- We now accept variants in VCF files with two alternative alleles (e.g. GT = 1/2). Those are included in the analysis as two separate heterozygous variants, which will be combined in pairs of compound heterozygous variants with the other heterozygous variants of that gene.
Release 2.3.0, May 2nd, 2023
VarCoPP2.0 is published in BMC Bioinformatics!
The model was slightly adapted in the revision process, with the main difference that it is a Balanced Random Forest of 400 instead of 100 trees. The updated version is now integrated in ORVAL.
All details can be found in the Documentation.
Release 2.2.0, January 25th, 2023
We made a change in the VCF parsing. Instead of automatically choosing the first alternative allele when multiple options are given (separated by commas), ORVAL will now use the GT field from the VCF file to pick the right ALT for each variant. See the Documentation section.
Release 2.1.0, January 11th, 2023
We made a change in the digenic combination creation step of the ORVAL pipeline.
Now, multiple heterozygous variants in the same gene are used together as heterozygous compound variants (combined with another gene) and not individually anymore.
This is to solve the problem of certain genes being over-represented in the results just because they contain heterozygous compound variants.
If a single gene contains more than two heterozygous variants, these are used in pairs of two as we always use two mutated alleles per gene. See the Documentation section.
Release 2.0.0, July 1st, 2022
This release features an improved version of the VarCoPP pathogenicity predictor, VarCoPP2.0. Read all about this new model in the Documentation section. You can use ORVAL in the same way as before this update. The implications are minor and include:
- New default variant filtering settings: MAF ≤ 0.035 instead of 0.03, and synonymous variants are removed if they are further than 195 nt (instead of 7 nt) away from the exon edge;
- New annotation features are used by VarCoPP2.0 and shown in the results. Notably, gene pairs are now ordered based on RVIS instead of GDI;
- The predictor is a single model with a single score output, as opposed to the ensemble of 500 Random Forest models in the first version of VarCoPP. Some of the result visualisations are adapted accordingly;
- The provided confidence zones are more stringent: 99% and 99.9%;
- ... More accurate and significantly faster predictions!
Release 1.1.1, April 13, 2022
- Added a job summary at top of the results page;
- Added statistics above S-plot in results (Digenic Predictions);
- Adopted ExAC to gnomAD integration;
- Upgraded technical dependencies.
Release 1.1, April 28, 2021
- Variant annotation database upgraded to use up-to-date versions of sources databases (cf. table below);
- Issue of missing splicing and intronic variants resolved;
- Retraining of VarCoPP and DE Predictor with the current annotations;
- Adaptation of VarCoPP confidence zones.
Release 1.0.1, June 6, 2020
- VCF parser improved to accept GT field for multiple alternate alleles. More information here: Documentation > VCF File
Current annotation database versions
|Resource||Version in GRCh37/hg19 database||Version in GRCh38/hg38 database|
|CADD||v1.6 (GRCh37/hg19)||v1.6 (GRCh38/hg38)|
|Ensembl||v101 (GRCh37/hg19)||v101 (GRCh38/hg38)|
|HGC (Biological distance)*||v12.2015||/|
|STRING DB (Biological distance)*||/||v11.5|
|BOCK (KG distance)||v1.0||v1.0|
* We re-calculated the Biological distance according to the method of the original paper (PMID: 24694260) in order to use the latest version of the STRING DB.