Posts

FAQ

How was the covariance data calculated?

2021-07-21 PredictDB Team
𝔽or the GTEx models, we calculated the covariance of the SNP dosages within our training (GTEx) samples. For the DGN model, we used public genotype data from 1000 Genomes to calculate the covariance of the SNP dosages. Read more →
FAQ

How was the expression data preprocessed for training the prediction models?

2021-07-21 PredictDB Team
𝔸: We used the expression pre-processed by the GTEx analysis working group. Covariates are the same as used in the eQTL analysis. GTEx processing pipeline is here(https://github.com/broadinstitute/gtex-pipeline/), and Docker image is here(https://hub.docker.com/r/broadinstitute/gtex_eqtl/). -[ ] … Read more →
FAQ

Q: Are performance measures of prediction available?

2021-07-21 PredictDB Team
𝔽or elastic net models, they can be found in the extra table in each sqlite db. The mashr-based models which use fine-mapping for variable selection, the performance measures are not available. You can use the elastic net performance as a proxy. Read more →
FAQ

Q: How many samples were used to train each model?

2021-07-21 PredictDB Team
𝕋his can be found in the sample_info table in each sqlite db. -[ ] TODO check that this info is available in the newer models. Read more →

tempo

2021-07-21 PredictDB Team
FAQ

There is an awful lot of models! Which ones should I use?

2021-07-21 PredictDB Team
𝔸1: GTEx v8 MASHR-based models are parsimonious and have been shown to yield more reliable set of putative causal variants. These are the best option. However, they require GWAS preprocessing on older GWAS as detailed here. A2: GTEx v8 UTMOST models are a robust but with lower performance. These are … Read more →
FAQ

Unexpected error: No objects to concatenate

2021-07-21 Festus
𝕋his error raise when the multixcan tool does get gwas files which should be concatenated. The error is common when you are provoding multiple gwas files in a directory. The common mistake that brings the issue rises from; […] –gwas_file_pattern “(.*).txt” #selects all gwas … Read more →
FAQ

Unexpected error: nothing to repeat at position 0"

2021-07-21 Festus
𝕋his error happens when the data bases cannot be loaded into the tool and the main cause might be; […] –models_name_pattern “en_(.*).db” –models_name_filter “en_Brain_(.*).db” Read more →
FAQ

What do i do with these models?

2021-07-21 PredictDB Team
𝕋hese models allow prediction of gene expression or alternative splicing on a GWAS study. The predicted levels can be associated to a complex trait such as a disease’s susceptibility. PrediXcan implementation uses these models and individual-level data. MetaXcan repository contains an … Read more →
FAQ

What does “TW_” mean in the GTEx v6 file names?

2021-07-21 PredictDB Team
𝕋his is an abbreviation for “Tissue-Wide.” We have historically created a Cross-Tissue model, which is a measure of the expression common across all tissues, and Tissue-Specific models (TS), which is modeled as an orthogonal component to the shared cross tissue component. Details can be found … Read more →