altos auto chemistry update

assets/protocols.json

@@ -1,7 +1,7 @@
 {
     "alevin": {
         "10XV1": {
-            "protocol": "10xv1",
+            "protocol": "1{b[14]u[10]x:}2{r:}",
             "whitelist": "assets/whitelist/10x_V1_barcode_whitelist.txt.gz"
         },
         "10XV2": {
@@ -13,7 +13,7 @@
             "whitelist": "assets/whitelist/10x_V3_barcode_whitelist.txt.gz"
         },
         "10XV4": {
-            "protocol": "10xv4",
+            "protocol": "1{b[16]u[12]x:}2{r:}",
             "whitelist": "assets/whitelist/10x_V4_barcode_whitelist.txt.gz"
         },
         "dropseq": {

conf/modules.config

@@ -59,8 +59,18 @@ process {
             enabled: false
         ]
     }
+    withName: AUTO_DETECT_PROTOCOL {
+        publishDir = [
+            path: { "${params.outdir}/pipeline_info" },
+            mode: params.publish_dir_mode,
+            saveAs: { filename -> filename.equals('versions.yml') ? null : filename },
+            enabled: false
+        ]
+    }
 }
 
+
+
 if(params.aligner == "cellranger") {
     process {
         withName: CELLRANGER_MKGTF {

modules/local/auto_detect_protocol.nf

@@ -0,0 +1,95 @@
+
+process AUTO_DETECT_PROTOCOL {
+    tag "$meta.id"
+    label 'process_single'
+
+    conda 'conda-forge::jq=1.6'
+    container "${ workflow.containerEngine == 'singularity' && !task.ext.singularity_pull_docker_container ?
+        'https://depot.galaxyproject.org/singularity/jq:1.6' :
+        'biocontainers/jq:1.6' }"
+
+    input:
+    // the first FastQ file in `reads` is expected to contain the cell barcodes
+    tuple val(meta), path(reads)
+    val aligner
+    path protocol_json
+    path barcode_whitelist
+
+    output:
+    tuple val(meta), path(reads), emit: ch_fastq
+    env PROTOCOL, emit: protocol
+    env EXTRA_ARGS, emit: extra_args
+    path "*.txt.gz", emit: whitelist
+    path "versions.yml", emit: versions
+
+    when:
+    task.ext.when == null || task.ext.when
+
+    script:
+    """
+
+    # convert protocols.json to table
+    TABLE=\$(
+        jq -r '
+            ."$aligner" |
+            to_entries[] |
+            "\\(.key)\\t\\(.value.protocol//"")\\t\\(.value.whitelist//"")\\t\\(.value.extra_args//"")"
+        ' "${protocol_json}"
+    )
+
+    # iterate over all protocols defined for the selected aligner
+    MATCHING_FRACTIONS=\$(cut -f1 <<<"\$TABLE" | while read KEY; do
+        # uncompress whitelist
+        WHITELIST=\$(grep -w "^\$KEY" <<<"\$TABLE" | cut -f3)
+        [ -n "\$WHITELIST" ] || continue # skip protocols without whitelist
+        WHITELIST_FILE=\$(basename "\$WHITELIST")
+        
+        gzip -dcf "\$WHITELIST_FILE" > barcodes
+
+        # subsample the FastQ file
+        gzip -dcf "${reads[0]}" |
+        awk 'FNR % 4 == 2' | # extract the read sequence from FastQ
+        head -n 100000 > reads || true # the first 100k reads should suffice
+
+        # extract the barcodes from the FastQ reads and count how many are valid barcodes
+        awk -v KEY="\$KEY" -v OFS='\\t' '
+            { \$0 = substr(\$0, 1, 14) } # the barcode is in the first 14 bases; 10X V2/3 barcodes are trimmed
+            FILENAME == "barcodes" { barcodes[\$0] } # cache barcodes in memory
+            FILENAME == "reads" && \$0 in barcodes { count++ } # count matches for each chemistry
+            END { print KEY, count/FNR } # output fraction of matching barcodes for each chemitry
+        ' barcodes reads
+
+    done | sort -k2,2gr)
+
+    # only trust the auto-detection if exactly one protocol matches
+    echo -e "These were the fractions of matching barcodes by protocol:\\n\$MATCHING_FRACTIONS"
+    MATCHING_PROTOCOLS_COUNT=\$(awk '\$2>=0.7' <<<"\$MATCHING_FRACTIONS" | wc -l)
+    if [ \$MATCHING_PROTOCOLS_COUNT -ne 1 ]; then
+         echo "ERROR: Found \$MATCHING_PROTOCOLS_COUNT matching protocols."
+         exit 1
+    fi
+    KEY=\$(cut -f1 <<<"\$MATCHING_FRACTIONS" | head -n1)
+
+    # extract attributes of chosen protocol
+    PROTOCOL=\$(grep -w "^\$KEY" <<<"\$TABLE" | cut -f2)
+    WHITELIST_PATH=\$(grep -w "^\$KEY" <<<"\$TABLE" | cut -f3)
+    WHITELIST=\$(basename "\$WHITELIST_PATH")
+    
+    # Remove all other whitelist files
+    for file in \$PWD/*.txt.gz; do
+        FILE_NAME=\$(basename "\$file")
+        [ "\$FILE_NAME" != "\$WHITELIST" ] && rm "\$FILE_NAME"
+    done
+
+    # Copy the chosen whitelist file
+    cp "\$WHITELIST" "whitelist.txt.gz"
+
+
+    EXTRA_ARGS=\$(grep -w "^\$KEY" <<<"\$TABLE" | cut -f4)
+    echo \$PWD/*.txt.gz
+    cat <<-END_VERSIONS > versions.yml
+    "${task.process}":
+        jq: \$(jq --version | cut -d- -f2)
+    END_VERSIONS
+    """
+}

modules/local/simpleaf_index.nf

@@ -33,6 +33,7 @@ process SIMPLEAF_INDEX {
     simpleaf set-paths
 
     # run simpleaf index
+
     simpleaf \\
         index \\
         --threads $task.cpus \\