Give me genetic information and trait information, and I can perform Transcriptome-Wide Association Analysis using the UKBB 300,000 European genotype data.
However, this project does not integrate any resources (such as UKBB data, or various method libraries), as they are simply too large. The purpose of writing this project is purely for orchestration, because manually processing and running each one is far too troublesome.
To run it, you first need to replace the constants in src/main/resources/env.properties with your own.
There are also some constants in UKB.java that need to be replaced. Finally, compile it into a jar package.
Please enter the absolute file path!
--genes [input_file]
【file content】
-- a gene each line，format
GeneSymbol,NCBIGene,Ensembl,chr,start,end
-- such as
TRIM24,8805,ENSG00000122779,7,138145004,138274741
--traits [trait1,trait2,...,traitn]
【file content】【example】
type:[Q|B]
threshold:30
dataId:21001,23104
ICD10:K70.0,K76.0
self-report:high cholesterol
type: Trait type, continuous (Q) or binary (B)
UK BioBank currently provides three sources of traits:
- dataId: The threshold matches the dataId, and the threshold is used to split the trait into cases and controls.
- ICD10
- self-report
It is not necessary to fill in all; choose based on the specific trait.
The above applies to automatic extraction of simple traits. Another option is provided below, where users need to manually extract traits.
--plinks [trait1,trait2,...,traitn]
You can directly input the bed/bim/fam (without extension) of already extracted traits.
Both methods can be used together.
--sex
Whether to consider sex, default is not to consider.
The following analysis methods are currently supported. Related resources can be downloaded from the links provided.
--magmaExecutable, Gene locations, build 37, European reference data--pascalDownload--fusionDownload--isoDownload--tfDownload--utmostDownload--spredDownload--smrDownload--ldscDownload
By default, all steps are run. You can use the following options to run only a specific step or steps, provided the preceding steps have succeeded.
--step1
Execute only preprocessing.
--step2
Run only the analysis.
--step3
Collect only the results.