POSTRE strategy to predict structural variants (SVs) pathogenicity is a cell-type-focused approach. This implies that POSTRE independently evaluates SV pathogenicity for the different cell types and tissues considered relevant to the phenotype of interest (details of all cell types used). For each of those cell types POSTRE considers and requires genomic data, such as, gene expression and enhancer maps. Evaluating an SV's impact on specific cell types is crucial for accurately identifying SV-associated changes in gene expression, especially those driven by enhancer-mediated alterations. 🧬
This cell-type-focused strategy, although essential to properly identify changes in gene-enhancer communication, may miss "obvious" pathogenic coding predictions if the relevant cell type for a disease is not included in POSTRE genomic database. For example, POSTRE only considers a gene as a Loss of Function (LOF) candidate in a given cell type if it's actively expressed on it. If a gene is inactive in all of the cell types currently in our database, it will never be identified as a pathogenic LOF candidate, even if it's truncated by an SV and known to be associated with a specific phenotype.
To address this limitation and prevent missing these clear coding alterations (in case you have not already evaluated them with the usage of other tools), we've introduced the Cell Type Agnostic Prediction option. This approach simply highlights as pathogenic: alterations in copy number and/or truncations of phenotypically relevant genes (e.g. associated in OMIM with the phenotype of interest) which are dosage sensitive. Thus, ignoring cell-type-specific genomic data.
When using the cell type agnostic prediction, it's important to be cautious with gene duplications. While the deletion of a gene will always lead to the loss of its expression wherever it gets activated, a duplication might only cause an increase/ectopic gene expression in specific cases, such as when enhancer adoption occurs. For example, as described in Trivellin et al 2025, isolated GPR101 duplications will not cause its ectopic expression and trigger X-linked acrogigantism unless enhancers from the neigbhor TAD are included on the duplication. Therefore, pathogenic gain of function predictions in this context should be interpreted with care. ⚠️