From 1d0d08cb1e6d8bb421194e8e88d1547b764c77e1 Mon Sep 17 00:00:00 2001
From: skchronicles <kuhnsa3@gmail.com>
Date: Tue, 19 Sep 2023 14:14:49 -0400
Subject: [PATCH] Adding sqanti ML filter rule

---
 workflow/Snakefile        |  5 +++++
 workflow/rules/sqanti.smk | 30 ++++++++++++++++++++++++++++--
 2 files changed, 33 insertions(+), 2 deletions(-)

diff --git a/workflow/Snakefile b/workflow/Snakefile
index 6a6fcf4..58d1fc7 100644
--- a/workflow/Snakefile
+++ b/workflow/Snakefile
@@ -264,6 +264,10 @@ rule all:
         # Characterize and annotate collapsed transcripts from flair
         # @imported from `rule sqanti_qc` in rules/sqanti.smk
         join(workpath, "project", "counts", "novel", "sqanti.isoforms_classification.txt"),
+        # SQANTI3 ML Filter, known/novel isoform quantification
+        # Filters characterized transcripts from Sqanti QC
+        # @imported from `rule sqanti_ml_filter` in rules/sqanti.smk
+        join(workpath, "project", "counts", "novel", "sqanti.isoforms_MLresult_classification.txt"),
         # Nanopolish polyA tail length estimation,
         # needs index to map basecalled reads to
         # raw signal from the ONT sequencer
@@ -332,3 +336,4 @@ include: join("rules", "editing_a-to-i.smk")
 include: join("rules", "polya.smk")
 include: join("rules", "diffexp.smk")
 include: join("rules", "altsplice.smk")
+include: join("rules", "hooks.smk")
\ No newline at end of file
diff --git a/workflow/rules/sqanti.smk b/workflow/rules/sqanti.smk
index e68eb0e..c062045 100644
--- a/workflow/rules/sqanti.smk
+++ b/workflow/rules/sqanti.smk
@@ -77,9 +77,35 @@ rule sqanti_ml_filter:
     https://github.com/ConesaLab/SQANTI3/wiki/
     Github: https://github.com/ConesaLab/SQANTI3
     @Input:
-        Sqanti Classification file (TSV) 
-    @Output:
+        Sqanti Classification file (TSV),
+        Corrected Annotation (GTF),
+        Corrected Transcriptome (FASTA)
+    @Input:
         ML Filtered Sqanti Classification file (TSV),
         ML Filtered Corrected Annotation (GTF),
         ML Filtered Corrected Transcriptome (FASTA)
     """
+    input:
+        txt = join(workpath, "project", "counts", "novel", "sqanti.isoforms_classification.txt"),
+        fa  = join(workpath, "project", "counts", "novel", "sqanti.isoforms_corrected.fasta"),
+        gtf = join(workpath, "project", "counts", "novel", "sqanti.isoforms_corrected.gtf"),
+    output:
+        txt = join(workpath, "project", "counts", "novel", "sqanti.isoforms_MLresult_classification.txt"),
+        fa  = join(workpath, "project", "counts", "novel", "sqanti.isoforms.filtered.fasta"),
+        gtf = join(workpath, "project", "counts", "novel", "sqanti.isoforms.filtered.gtf"),
+    params:
+        rname  = "sqanti_ml_filter",
+        prefix = "sqanti.isoforms",
+        outdir      = join(workpath, "project", "counts", "novel"),
+    container: depending(config['images']['sqanti3'], use_singularity),
+    threads: int(allocated("threads", "sqanti_ml_filter", cluster)),
+    shell: """
+    # Applies ML filter on selected
+    # SQANTI3 QC classification file
+    sqanti3_filter.py ML \\
+        --output {params.prefix} \\
+        --dir {params.outdir} \\
+        --gtf {input.gtf} \\
+        --isoforms {input.fa} \\
+        {input.txt}
+    """
\ No newline at end of file