Solve-RD RNA-seq paper

This repository is intended to allow reproducing the results of the Solve-RD RNA-seq flagship paper, where 20 families affected by a rare disorder where diagnosed.

The analysis is composed of 3 independent omics layers:

1. RNA: Expression and splicing outlier detection on RNA-seq data from 4 different tissues.
2. DNA: variant annotation and prioritization.
3. Phenotype: computing of HPO scores. Refer to the manuscript for a description of the Methods.

Most of the RNA-seq and DNA samples are deposited in the European Genome-Phenome Archive (EGA), where they can be requested to download. The request will be reviewed by the Solve-RD Data Access Committee. The Solve-RD EGA study number is EGAS00001003851.

RNA-seq

RNA-seq data was collected from four tissues: whole blood, muscle, PBMCs and fibroblasts. Aberrant expression and splicing analysis were performed independently for each tissue using DROP, v1.3.0.

To reach statistical significance, some Solve-RD samples were merged with external data from the same tissue and similar sequencing protocol. All the Solve-RD RNA-seq data is are deposited in EGA, with dataset number EGAD00001016121. The count matrices of external data are shared in Zenodo. See the rna folder for more details.

Disclaimer

The results may vary due to various factors including:

• Software updates in Bioconductor, OUTRIDER, FRASER, Python, Java, among others
• Using OHT to find the optimal encoding dimension
• Due to consent reasons, the external count matrices only include junctions for which both start and end exist in the GENCODE annotation file. The splicing analyses will be affected by this.

However, extreme outliers should remain.

DNA

Whole-exome or whole-genome sequencing was also performed for the individuals with RNA-seq samples. VCF files for some of them are also available in EGA. Rare SNVs and indels were extracted to them and combined with the RNA-seq resutls for interpretation.

See the dna folder for more details.

HPO scores

Each affected individual had their clinical phenotype collected in the form of HPO (Human Phenotype Ontology) terms. See the hpo folder how to compute them.

Sample annotation

The RNA & DNA sample annotation tables are available for download in Zenodo. The sample annotation of the count matrices are available in their corresponding Zenodo folder.

Contact

See the corresponding authors of the published study.