v1.3.1

This version fixes a TWAS sign swap bug introduced in v1.0.4 (2025-03-02).

plink2R had been replaced with snpStats to avoid manual install step. But snpStats loads genotype dosages in terms of A2 allele rather than the A1 allele as plink does. This led to top1 and enet model types having flipped weights since they used the loaded genotypes, while the other model types were computed via subcommands. TWAS associations and phenotype predictions based on the models had flipped signs. Now, loaded dosages are subtracted from 2 to flip back to the A1-based representation expected by downstream applications.

v1.3.0

LaDDR integration. As an optional stage after computing RNA phenotypes, LaDDR can be run to produce latent, data-driven RNA phenotypes. The knowledge-driven phenotypes are regressed out of the LaDDR input data to produce residual latent phenotypes that can be included in genetic analyses such as cross-modality xQTL mapping and xTWAS.

v1.2.0

Expands compatibility with GTFs from sources beyond Ensembl. Phenotyping should now also work with Gencode and RefSeq GTF files.

The calculation of exonic to intronic read ratio has also changed, both to work with non-Ensembl GTF files and as a theoretical improvement:

• Instead of reporting exons found in at least a certain percentage of isoforms, report regions that are covered by exons in at least a certain percentage of isoforms. That allows, e.g., a region to be included if some isoforms have exons that extend beyond the region. Before, exon boundaries had to match exactly across isoforms after filtering by Ensembl-specific annotation sources.
• Intronic regions are defined as regions within the gene (as defined by exons) that contain no exons in any isoform.

v1.1.1

• Alternative TSS and alternative polyA phenotypes can now be produced for lncRNA genes. Previously they were inadvertently omitted by txrevise.
• Other bug fixes

v1.1.0

Adds a seventh RNA phenotype modality, RNA editing ratio quantification.

v1.0.5

• Updated phenotype ID format
• genome_size is no longer required in config
• Various other fixes

v1.0.4

• Allow mixed single and paired end
• Improve handling of TWAS errors
• Replace plink2R and VariantAnnotation with snpStats
• Rename project directories
• Rename cross-modality files
• Script to prepare phenotypes for Pheast

v1.0.3

Updated slurm resources and fixed dependencies. Added TWAS association testing template.

v1.0.2

Various bug fixes related to dependencies and data formats. Added trans-xQTL mapping.

v1.0.1

Various bug fixes for single-end reads and other runtime issues.