Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Move method information out of the meta, following Jon's code review … #375

Open
wants to merge 3 commits into
base: dev-ratio
Choose a base branch
from
Open

Move method information out of the meta, following Jon's code review … #375

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
3 commits
Select commit Hold shift + click to select a range
fffab93
Move method information out of the meta, following Jon's code review …
Dec 5, 2024
5b228a5
fix linting
Dec 5, 2024
df06636
minor cleaning
Dec 18, 2024
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
13 changes: 5 additions & 8 deletions subworkflows/local/correlation/main.nf
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -5,7 +5,7 @@ include {PROPR_PROPR as PROPR} from "../../../modules/local/propr/propr/main.nf"

workflow CORRELATION {
take:
ch_counts // [ meta, counts] with meta keys: method, args_cor
ch_counts // [[meta_input], counts, analysis method]

main:

Expand All @@ -15,22 +15,19 @@ workflow CORRELATION {
ch_versions = Channel.empty()

// branch tools to select the correct correlation analysis method
ch_counts
.branch {
propr: it[0]["method"] == "propr"
inputs = ch_counts
.map { meta, abundance, analysis_method ->
[ meta + ['method': analysis_method ], abundance]
}
.set { ch_counts }

// ----------------------------------------------------
// Perform correlation analysis with propr
// ----------------------------------------------------
PROPR(ch_counts.propr.unique())
PROPR( inputs.filter{it[0].method == 'propr'})
ch_matrix = PROPR.out.matrix.mix(ch_matrix)
ch_adjacency = PROPR.out.adjacency.mix(ch_adjacency)
ch_versions = ch_versions.mix(PROPR.out.versions)

// TODO: divide propr module into cor, propr, pcor, pcorbshrink, etc.

emit:
matrix = ch_matrix // channel: [ csv ]
adjacency = ch_adjacency // channel: [ csv ]
Expand Down
88 changes: 23 additions & 65 deletions subworkflows/local/differential/main.nf
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -10,11 +10,12 @@ include { FILTER_DIFFTABLE as FILTER_DIFFTABLE_DESEQ2 } from '../../../modules/l

workflow DIFFERENTIAL {
take:
ch_counts // [ meta_exp, counts ] with meta keys: method, args_diff
ch_input // [[meta_input], counts, analysis method]

ch_samplesheet // [ meta_exp, samplesheet ]
ch_contrasts // [ meta_contrast, contrast_variable, reference, target ]
ch_transcript_lengths
ch_control_features
ch_transcript_lengths // [ meta_exp, transcript_lengths]
ch_control_features // [ meta_exp, control_features]

main:

Expand All @@ -26,14 +27,17 @@ workflow DIFFERENTIAL {
ch_adjacency = Channel.empty()
ch_versions = Channel.empty()

// branch tools to select the correct differential analysis method
ch_counts
.branch {
propd: it[0]["method"] == "propd"
deseq2: it[0]["method"] == "deseq2"
limma: it[0]["method"] == "limma"
// We need to cross the things we're iterating
inputs = ch_input
.combine(ch_samplesheet)
.filter{ meta_input, abundance, analysis_method, meta_exp, samplesheet -> meta_input.subMap(meta_exp.keySet()) == meta_exp }
.combine(ch_contrasts)
.multiMap { meta_input, abundance, analysis_method, meta_exp, samplesheet, meta_contrast, variable, reference, target ->
def meta = meta_input.clone() + meta_contrast.clone() + ['method': analysis_method ] // ToDo: If modules are updated to combine their input metas, then this merge could be removed
samples_and_matrix: [ meta, samplesheet, abundance]
contrasts: [ meta, variable, reference, target]
propd_input: [ meta + ['contrast': meta_contrast.id, 'method': analysis_method], abundance, samplesheet, variable, reference, target ] // ToDo: remove this, once propd is updated and uses standardized inputs
}
.set { ch_counts }

// ----------------------------------------------------
// Perform differential analysis with propd
Expand All @@ -42,18 +46,7 @@ workflow DIFFERENTIAL {
// TODO propd currently don't support blocking, so we should not run propd with same contrast_variable, reference and target,
// but different blocking variable, since it will simply run the same analysis again.

ch_counts.propd
.combine(ch_samplesheet)
.filter{ meta_counts, counts, meta_samplesheet, samplesheet -> meta_counts.subMap(meta_samplesheet.keySet()) == meta_samplesheet }
.combine(ch_contrasts)
.map {
meta_data, counts, meta_samplesheet, samplesheet, meta_contrast, contrast_variable, reference, target ->
def meta = meta_data.clone() + ['contrast': meta_contrast.id]
return [ meta, counts, samplesheet, contrast_variable, reference, target ]
}
.set { ch_propd }

PROPD(ch_propd.unique())
PROPD( inputs.propd_input.filter{it[0].method == 'propd'})
ch_results_pairwise = PROPD.out.results.mix(ch_results_pairwise)
ch_results_pairwise_filtered = PROPD.out.results_filtered.mix(ch_results_pairwise_filtered)
ch_results_genewise = PROPD.out.connectivity.mix(ch_results_genewise)
Expand All @@ -65,29 +58,17 @@ workflow DIFFERENTIAL {
// Perform differential analysis with DESeq2
// ----------------------------------------------------

ch_counts.deseq2
.combine(ch_samplesheet)
.filter{ meta_counts, counts, meta_samplesheet, samplesheet -> meta_counts.subMap(meta_samplesheet.keySet()) == meta_samplesheet }
.combine(ch_contrasts)
.multiMap {
meta_data, counts, meta_samplesheet, samplesheet, meta_contrast, contrast_variable, reference, target ->
def meta = meta_data.clone() + meta_contrast.clone()
contrast: [ meta, contrast_variable, reference, target ]
samples_and_matrix: [ meta, samplesheet, counts ]
}
.set { ch_deseq2 }

// do we need this process DESEQ2_NORM?
DESEQ2_NORM (
ch_deseq2.contrast.first(),
ch_deseq2.samples_and_matrix,
inputs.contrasts.filter{it[0].method == 'deseq2'}.first(),
inputs.samples_and_matrix.filter{it[0].method == 'deseq2'},
ch_control_features,
ch_transcript_lengths
)

DESEQ2_DIFFERENTIAL (
ch_deseq2.contrast,
ch_deseq2.samples_and_matrix,
inputs.contrasts.filter{it[0].method == 'deseq2'},
inputs.samples_and_matrix.filter{it[0].method == 'deseq2'},
ch_control_features,
ch_transcript_lengths
)
Expand All @@ -97,16 +78,6 @@ workflow DIFFERENTIAL {
ch_model_deseq2 = DESEQ2_DIFFERENTIAL.out.model
ch_versions = DESEQ2_DIFFERENTIAL.out.versions.mix(ch_versions)

ch_processed_matrices = ch_norm_deseq2
if ('rlog' in params.deseq2_vs_method){
ch_processed_matrices = ch_processed_matrices.join(DESEQ2_NORM.out.rlog_counts)
}
if ('vst' in params.deseq2_vs_method){
ch_processed_matrices = ch_processed_matrices.join(DESEQ2_NORM.out.vst_counts)
}
ch_processed_matrices = ch_processed_matrices
.map{ it.tail() }

// TODO modify the module to accept these parameters as meta/ext.args in the same way how propd does
ch_logfc_deseq2 = Channel.value([ "log2FoldChange", params.differential_min_fold_change ])
ch_padj_deseq2 = Channel.value([ "padj", params.differential_max_qval ])
Expand All @@ -125,24 +96,11 @@ workflow DIFFERENTIAL {
// Perform differential analysis with limma
// ----------------------------------------------------

// combine the input channels with the tools information
// in this way, limma will only be run if the user have specified it, as informed by ch_tools
ch_counts.limma
.combine(ch_samplesheet)
.filter{ meta_counts, counts, meta_samplesheet, samplesheet -> meta_counts.subMap(meta_samplesheet.keySet()) == meta_samplesheet }
.combine(ch_contrasts)
.unique()
.multiMap {
meta_data, counts, meta_samplesheet, samplesheet, meta_contrast, contrast_variable, reference, target ->
def meta = meta_data.clone() + meta_contrast.clone()
input1: [ meta, contrast_variable, reference, target ]
input2: [ meta, samplesheet, counts ]
}
.set { ch_limma }


// run limma
LIMMA_DIFFERENTIAL(ch_limma.input1, ch_limma.input2)
LIMMA_DIFFERENTIAL(
inputs.contrasts.filter{it[0].method == 'limma'},
inputs.samples_and_matrix.filter { it[0].method == 'limma' }
)
ch_versions = LIMMA_DIFFERENTIAL.out.versions.mix(ch_versions)

// filter results
Expand Down
30 changes: 14 additions & 16 deletions subworkflows/local/enrichment/main.nf
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -12,10 +12,10 @@ include { TABULAR_TO_GSEA_CHIP } from '../../../modules/local/tabular_to_gsea_ch

workflow ENRICHMENT {
take:
ch_counts // [ meta, counts] with meta keys: method, args_cor
ch_results_genewise // [ meta, results] with meta keys: method, args_cor
ch_results_genewise_filtered // [ meta, results] with meta keys: method, args_cor
ch_adjacency // [ meta, adj_matrix] with meta keys: method, args_cor
ch_counts // [[meta_input], counts, analysis method]
ch_results_genewise // [[meta_results], results, analysis method]
ch_results_genewise_filtered // [[meta_results], results, analysis method]
ch_adjacency // [[meta_adj_matrix], adj_matrix, analysis method]
ch_contrasts
ch_samplesheet
ch_featuresheet
Expand All @@ -31,14 +31,14 @@ workflow ENRICHMENT {
ch_versions = Channel.empty()

ch_counts
.branch {
gsea: it[0]["method"] == "gsea"
.map { meta, abundance, analysis_method ->
[ meta + ['method': analysis_method ], abundance]
}
.set { ch_counts }

ch_adjacency
.branch {
grea: it[0]["method"] == "grea"
.map { meta, adj_matrix, analysis_method ->
[ meta + ['method': analysis_method ], adj_matrix]
}
.set { ch_adjacency }

Expand All @@ -56,7 +56,7 @@ workflow ENRICHMENT {
// Perform enrichment analysis with GREA
// ----------------------------------------------------

GREA(ch_adjacency.grea.unique(), ch_gmt.collect())
GREA(ch_adjacency.filter{it[0].method == 'grea'}, ch_gmt.collect())
ch_enriched = ch_enriched.mix(GREA.out.results)
ch_versions = ch_versions.mix(GREA.out.versions)

Expand All @@ -68,7 +68,7 @@ workflow ENRICHMENT {
// input, and process the sample sheet to generate class definitions
// (CLS) for the variable used in each contrast

CUSTOM_TABULARTOGSEAGCT ( ch_counts )
CUSTOM_TABULARTOGSEAGCT ( ch_counts.filter{it[0].method == 'gsea'} )

// TODO: update CUSTOM_TABULARTOGSEACLS for value channel input per new
// guidlines (rather than meta usage employed here)
Expand All @@ -94,7 +94,6 @@ workflow ENRICHMENT {
.map{ tuple(it[1], it[0], it[2]) }
.combine( ch_gmt.map { meta, gmt -> gmt } )

println("__"+TABULAR_TO_GSEA_CHIP.out.chip)
GSEA_GSEA(
ch_gsea_inputs,
ch_gsea_inputs.map{ tuple(it[0].reference, it[0].target) }, // *
Expand Down Expand Up @@ -130,21 +129,20 @@ workflow ENRICHMENT {
ch_background = []
} else if (params.gprofiler2_background_file == "auto") {
// If auto, use input matrix as background
ch_background = ch_counts.map { meta, counts -> counts }
ch_background = ch_counts.filter{it[0].method == 'gprofiler2'}.map { meta, counts -> counts }
} else {
ch_background = Channel.from(file(params.gprofiler2_background_file, checkIfExists: true))
}

// rearrange channel for GPROFILER2_GOST process
ch_gmt = ch_gmt.map { meta, gmt -> gmt }

ch_results_genewise_filtered
.branch {
grea: it[0]["method"] == "gprofiler2"
.map { meta, results, analysis_method ->
[ meta + ['method': analysis_method ], results]
}
.set { ch_results_genewise_filtered }

GPROFILER2_GOST(ch_results_genewise_filtered, ch_gmt, ch_background)
GPROFILER2_GOST(ch_results_genewise_filtered.filter{it[0].method == 'gprofiler2'}, ch_gmt, ch_background)
ch_versions = ch_versions.mix(GPROFILER2_GOST.out.versions)

emit:
Expand Down
6 changes: 3 additions & 3 deletions subworkflows/local/experimental/main.nf
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -9,11 +9,11 @@ def preprocess_subworkflow_output( ch_input, ch_tools_args, method_field_name) {
// add method arguments to channel meta
return ch_input
.combine(ch_tools_args)
.filter{ meta, input, pathway, arg_maps -> meta["pathway_name"] ? meta["pathway_name"] == pathway["pathway_name"] : true }
.filter{ meta, input, pathway, arg_map -> meta["pathway_name"] ? meta["pathway_name"] == pathway["pathway_name"] : true }
.map{ meta, input, pathway, arg_map ->
def meta_clone = meta.clone() + pathway + arg_map.clone()
meta_clone["method"] = meta_clone.remove(method_field_name)
return [meta_clone, input]
def method = meta_clone.remove(method_field_name)
return [meta_clone, input, method]
}
}

Expand Down
Loading