Added drep swf doe large catalogues

subworkflows/drep_large_catalogue_swf.nf

@@ -0,0 +1,101 @@
+/*
+ * De-replicate
+ */
+
+include { DREP_CHUNKED } from '../modules/drep_chunked'
+include { DREP_RERUN } from '../modules/drep_chunked'
+include { COMBINE_CHUNKED_DREP } from '../modules/combine_chunked_drep'
+include { SPLIT_DREP_LARGE } from '../modules/split_drep_large'
+include { CLASSIFY_CLUSTERS } from '../modules/classify_clusters'
+include { MASH_COMPARE } from '../modules/mash_compare'
+
+workflow DREP_LARGE_SWF {
+ take:
+ genomes_directory
+ checkm_csv
+ extra_weight_table
+ main:
+
+ files = fileTree(dir: genomes_directory).filter { it.isFile() }
+
+ // split genomes from genomes_directory into chunks 25000 files each (probably doesnt work)
+ fileGroups = files.collect { file ->
+ groupTuple = fileGroups.size() > 0 ? fileGroups[-1] : null
+ if (groupTuple == null || groupTuple.size() == 25000) {
+ groupTuple = []
+ fileGroups << groupTuple
+ }
+ groupTuple << file
+ }
+
+ chunks = genomes_directory.groupTuple(fileGroups)
+
+ // run process on each chunk
+ DREP_CHUNKED(
+ chunks,
+ checkm_csv,
+ extra_weight_table)
+
+ // collect all species representative fasta files (emitted as dereplicated_genomes directories)
+ // THIS PROBABLY DOESNT WORK
+ drep_dirs = fileGroups.collect { subdir ->
+ path("DREP_CHUNKED.out.dereplicated_genomes", subdir)
+ }
+
+ all_dereplicated_genomes = drep_dirs.collectMany { resultDir ->
+ fileTree(resultDir).filter { it.isFile() }
+ }
+
+ // run drep on species representatives from all DREP_CHUNKED runs
+ DREP_RERUN(
+ chunks,
+ checkm_csv,
+ extra_weight_table)
+
+ // THIS NEEDS REAL CODE
+ all_cdb_files = gather all DREP_CHUNKED.out.cdb_csv into an array
+ all_sdb_files = gather all DREP_CHUNKED.out.sdb_csv into an array
+
+ // run a python script to combine the outputs
+ COMBINE_CHUNKED_DREP(
+ all_cbd_files,
+ all_sbd_files,
+ DREP_RERUN.out.cbd_csv
+ )
+
+ SPLIT_DREP_LARGE(
+ DREP.out.cdb_csv,
+ DREP.out.sdb_csv
+ )
+
+ CLASSIFY_CLUSTERS(
+ genomes_directory,
+ SPLIT_DREP.out.text_split
+ )
+
+ groupGenomes = { fna_file ->
+ def cluster = fna_file.parent.toString().tokenize("/")[-1]
+ return tuple(cluster, fna_file)
+ }
+
+ // Run mash on each group of fastas in many_genomes_fnas
+ // THESE NEED TO BE PUBLISHED
+ MASH_COMPARE(
+ CLASSIFY_CLUSTERS.out.many_genomes_fnas
+ )
+
+ // TODO: cat all mash outputs into one Mdb.csv file and add to emit list
+
+ many_genomes_fna_tuples = CLASSIFY_CLUSTERS.out.many_genomes_fnas | flatten | map(groupGenomes)
+ single_genomes_fna_tuples = CLASSIFY_CLUSTERS.out.one_genome_fnas | flatten | map(groupGenomes)
+
+ emit:
+ many_genomes_fna_tuples = many_genomes_fna_tuples
+ single_genomes_fna_tuples = single_genomes_fna_tuples
+ drep_split_text = SPLIT_DREP_LARGE.out.text_split
+ // mash splits need to be pulled together from each run of MASH_COMPARE
+ mash_splits = MASH_COMPARE.out.mash_split
+ drep_cdb_csv = DREP.out.cdb_csv
+ //drep_mdb_csv = DREP.out.mdb_csv
+ drep_sdb_csv = DREP.out.sdb_csv
+}