posts/h3a_bionet_worfklowhackathon.mdwn

   1 [[!meta title="H3ABioNet Hackathon (Workflows)"]]
   2
   3 I'm in Pretoria, South Africa at the
   4 [H3ABioNet](http://h3abionet.org/)
   5 [hackathon](http://h3abionet.org/17-h3abionet-courses/h3abionet-courses-upcoming/266-h3abionet-cloud-computing-hackathon)
   6 which is developing workflows for Illumina chip genotyping,
   7 imputation, 16S rRNA sequencing, and population structure/association
   8 testing. Currently, I'm working with the imputation stream and we're
   9 using [Nextflow](https://www.nextflow.io/) to deploy an
  10 [IMPUTE](https://mathgen.stats.ox.ac.uk/impute/impute_v2.html)-based
  11 imputation workflow with Docker and
  12 [NCSA's openstack-based cloud (Nebula)](https://wiki.ncsa.illinois.edu/display/NEBULA/Nebula+Home)
  13 underneath.
  14
  15 The OpenStack command line clients (`nova` and `cinder`) seem to be
  16 pretty usable to
  17 [automate bringing up a fleet of VMs](https://github.com/h3abionet/chipimputation/blob/master/openstack/generate_openstack)
  18 and the cloud-init package which is present in the images makes
  19 [configuring the images pretty simple](https://github.com/h3abionet/chipimputation/tree/master/openstack).
  20
  21 Now if I just knew of a better shared object store which was supported
  22 by Nextflow in OpenStack besides mounting an NFS share, things would
  23 be better.
  24
  25 You can follow our progress in our git repo:
  26 https://github.com/h3abionet/chipimputation
  27
  28
  29 [[!tag debian tech biology workflows]]