Home » AGAPS » AGAPS Tutorials » Pipeline Tutorial

Pipeline Tutorial

Pipeline Version 2.07

To run pipeline:

make pipeline executable file (you have to do it any time after saving in any Windows editor, e.g., notepad++):
```
chmod 755 pipeline.sh
```
clean the directory
```
./pipeline CHS REMOVE
```
run pipeline
```
./pipeline CHS
```
where you deal with CHS data.

Instead you can run:

LLFS
HRS
MESA
ARIC
FRAM_CARe
FRAM_SHARe

etc.

You need to use (or to prepare or edit) Input Files:

CHS.sas7bdat

Usually this file is ready and in the directory source_data. Just check that you have the file. This file contains information about individuals to be analyzed and variables represented by phenotype and covariates.

Main options you could change are:

QUALITYCONTROL="--mind 0.05 --hwe 0.001 --maf 0.05 --geno 0.05"
SKIPDATA=0
ANNOTATION=0
PCA=0

You can use SKIPDATA=1 if you would like to skip data preparation step.
You can use ANNOTATION=1 to prepare genetic data for analysis. Fist run of pipeline for a specific data should be with ANNOTATION=1.
Option PCA is still under construction. Please always use PCA=0.

You can also choose sex among

#SEX="MALE"
#SEX="FEMALE"
SEX="TOTAL"
#SEX="MALE FEMALE"

and methods for calculation of associations. Main approach is

UNIMETHOD="EMMAX ALLSAS,ALLSASG"

By default you analyze effects of SNP on mortality. You can change the outcome by changing lines after

if [ "${SKIPDATA}" == 0 ]; then

i.e., lines like this:

PHENO_FRAM_CARe="lefilled2"

and/or

OTHER_PHENO_LLFS="age_sdi LifeSpan_FU isdead_fu 95"

and/or covariates in

COVARIATES_LLFS="SexM yob_sdi"