Run data analysis jobs in Kubernetes.
Note: The code below is separate from the related functionality available in pipeline5 and platinum. Results are not guaranteed to be identical or even similar.
| Job name | Description |
|---|---|
| dna_align | Run bwa mem alignment of paired reads FASTQ. |
| rna_align | Run STAR alignment of paired reads RNA FASTQ. |
| non_umi_dedup | Deduplicate bam with sambamba markdup whilst ignoring UMI's. |
| umi_dedup | Deduplicate bam with UMI-Collapse by taking UMI's into account. |
| flagstat | Collect flagstat stats for bam with sambamba flagstat. |
| count_mapping_coords | Count unique mapping positions present in bam. |
You need to have the right credentials and a cluster (only need this step once).
gcloud container clusters get-credentials rerun-cluster --region europe-west4 --project hmf-crunch>All commands provide more complete documentation when run with incomplete or incorrect arguments. Some suggested commands to get you started:
./k8analysis
./k8analysis run
./k8analysis run dna_align
./k8analysis run non_umi_dedup
./k8analysis version
./k8analysis build
./k8analysis push
./k8analysis set_defaultThe FASTQ file names need to contain _R1_ or _R2_ to show whether they are read 1 or 2.
All read 1 FASTQ files need to have a corresponding read 2 FASTQ file and vice versa.
A read 1 FASTQ file corresponds to a read 2 FASTQ file when the only difference in their file names is the _R1_ vs _R2_.
kubectl get jobs | grep <your-job-name>
kubectl get pods | grep <your-job-name>
kubectl logs <your-pod-name>