  
This project ports the IBS/MDS/IBD functionality of PLink to Spark / ADAM.
Checkout our wiki
JDK 1.7
Scala 2.10.x
adam-ibs is composed of modules adam-ibs-core and adam-ibs-data.
adam-ibs-data define the used avro data model. This module needs to be builded in order to be used in adam-ibs-core module as a .jar dependence.
adam-ibs-core include logic and lci and uses adam-ibs-data jar dependency.
The project can be built by executing "mvn clean compile install" from the root folder or in two following steps :
- In adam-ibs-data folder : maven clean compile install
- You will obtain both adam-ibs-data-0.1.0.jar which is used as maven dependency in adam-ibs-core.
- In adam-ibs-code folder : maven clean compile install
- You will obtain 2 jar files :
- adam-ibs-core-0.1.0.jar -> jar file containing adam-ibs-core code
- adam-ibs-core-0.1.0-jar-with-dependencies.jar -> jar file containing adam-ibs-core code with all dependencies.
From adam-ibs-core folder :
Help :
adam-ibs-core$ java -jar target/adam-ibs-core-0.1.0-jar-with-dependencies.jar -h
21:01:07.849 [main] [INFO ] [c.e.m.c.Main$] : Begin with arguments : -h
--file <name> Specify .ped + .map filename prefix (default 'plink')
--genome Calculate IBS distances between all individuals [needs
--file and --out]
--make-bed Create a new binary fileset. Specify .ped and .map files [required --file and --out]
--out <name> Specify the output filename
--show-parquet Show Schema and data sample stored in a parquet file [required --file]
-h, --help <arg> Show help message
Execute with spark-submit :
spark-1.4.1/bin$ ./spark-submit --master local /home/ikizema/DEV/mgl804/adam-ibs/adam-ibs-core/target/adam-ibs-core-0.1.0-jar-with-dependencies.jar --file /home/ikizema/DEV/mgl804/adam-ibs/DATA/avro/test.makeBed.parquet --show-parquet
CLI parser is integrated to code.
-
View CLI options (--help) :
-
Exemple : Execute --make-bed on --file and get a parquet output file --out
