This routine calls one of four phases in a parallelized version of qtlnet.
parallel.qtlnet(phase, index = 1, …, dirpath = ".")Phase of parallelization as number 1 through 4. See details.
Index for phase. Used in phases 2 and 4, and for error codes saved
in RESULT.phase.index file.
Additional arguments for phases. See details.
Character string for directory were user can read and write files. When submitting to a cluster, this should remain the default.
See http://www.stat.wisc.edu/~yandell/sysgen/qtlnet for details of
implementation in progress. The plan is to run qtlnet via Condor
(https://research.cs.wisc.edu/htcondor/) to scale up to larger networks, say up to
100 nodes. Most important information is passed in files. Phase 1
imports arguments from the params.txt file, which must have
parse-able assignments to the arguments of
qtlnet:::qtlnet.phase1. This first phase produces file
Phase1.RData, which included objects used by all other phases.
Phase 1 also creates file groups.txt, which for each line has
begin and end indices for the parents that would result from a
call to parents.qtlnet. Phase 2 should be run the same
number of times as the number of lines in file groups.txt. Each
run produces a bicN.RData file containing BIC computations. These
computations are aggregated in Phase 3 to create Phase3.RData,
which contains the saved.scores used for
mcmc.qtlnet runs in Phase 4, which each produce an
mcmcN.RData file. The number of runs of Phase 4
is an argument nruns stored in the params.txt file
processed in Phase 1. Finally, Phase 5 aggregates the MCMC results from
multiple independent runs into one qtlnet object.
Chaibub Neto E, Keller MP, Attie AD, Yandell BS (2010) Causal Graphical Models in Systems Genetics: a unified framework for joint inference of causal network and genetic architecture for correlated phenotypes. Ann Appl Statist 4: 320-339. http://www.stat.wisc.edu/~yandell/doc/2010/92.AnnApplStat.pdf
# NOT RUN {
# }
# NOT RUN {
parallel.qtlnet("/u/y/a/yandell/public/html/sysgen/qtlnet/condor", 1)
# }
Run the code above in your browser using DataLab