Add Batch Controls Mode supporting fixed variant coordinates and customizable variant types, VAF, and indel sizes

pipeline.py

@@ -307,7 +307,9 @@ def generate_synthetic_bam(
     insert_size, insert_std, indel_interval, read_length=150,
     sequencing_mode="hybrid_capture",
     log_func=None, progress_func=None,
-    compress_fastq=False
+    compress_fastq=False,
+    variant_type="Mixed",
+    indel_length=0
 ):
     import pysam
     from collections import defaultdict
@@ -344,18 +346,26 @@ def generate_synthetic_bam(
                 ref_base = fasta.fetch(chrom, pos, pos + 1).upper()
             except Exception:
                 continue
-            if indel_interval > 0 and (variant_count + 1) % indel_interval == 0:
-                indel_len = random.randint(1, 4)
-                if random.choice(["INS", "DEL"]) == "INS":
-                    ins_bases = "".join(random.choices(["A", "C", "G", "T"], k=indel_len))
-                    ref_seq, alt_seq = ref_base, ref_base + ins_bases
+            # Determine variant type for this locus
+            current_type = variant_type
+            if variant_type == "Mixed":
+                if indel_interval > 0 and (variant_count + 1) % indel_interval == 0:
+                    current_type = random.choice(["Insertion", "Deletion"])
                 else:
-                    try:
-                        ref_seq = fasta.fetch(chrom, pos, pos + indel_len + 1).upper()
-                    except Exception:
-                        ref_seq = ref_base
-                    alt_seq = ref_base
-            else:
+                    current_type = "SNV"
+
+            if current_type == "Insertion":
+                curr_len = indel_length if indel_length > 0 else random.randint(1, 4)
+                ins_bases = "".join(random.choices(["A", "C", "G", "T"], k=curr_len))
+                ref_seq, alt_seq = ref_base, ref_base + ins_bases
+            elif current_type == "Deletion":
+                curr_len = indel_length if indel_length > 0 else random.randint(1, 4)
+                try:
+                    ref_seq = fasta.fetch(chrom, pos, pos + curr_len + 1).upper()
+                except Exception:
+                    ref_seq = ref_base
+                alt_seq = ref_base
+            else:  # SNV
                 alt_bases = [b for b in ["A", "C", "G", "T"] if b != ref_base]
                 alt_seq = random.choice(alt_bases) if alt_bases else "A"
                 ref_seq = ref_base

src/pipeline.py

@@ -307,7 +307,9 @@ def generate_synthetic_bam(
     insert_size, insert_std, indel_interval, read_length=150,
     sequencing_mode="hybrid_capture",
     log_func=None, progress_func=None,
-    compress_fastq=False
+    compress_fastq=False,
+    variant_type="Mixed",
+    indel_length=0
 ):
     import pysam
     from collections import defaultdict
@@ -344,18 +346,26 @@ def generate_synthetic_bam(
                 ref_base = fasta.fetch(chrom, pos, pos + 1).upper()
             except Exception:
                 continue
-            if indel_interval > 0 and (variant_count + 1) % indel_interval == 0:
-                indel_len = random.randint(1, 4)
-                if random.choice(["INS", "DEL"]) == "INS":
-                    ins_bases = "".join(random.choices(["A", "C", "G", "T"], k=indel_len))
-                    ref_seq, alt_seq = ref_base, ref_base + ins_bases
+            # Determine variant type for this locus
+            current_type = variant_type
+            if variant_type == "Mixed":
+                if indel_interval > 0 and (variant_count + 1) % indel_interval == 0:
+                    current_type = random.choice(["Insertion", "Deletion"])
                 else:
-                    try:
-                        ref_seq = fasta.fetch(chrom, pos, pos + indel_len + 1).upper()
-                    except Exception:
-                        ref_seq = ref_base
-                    alt_seq = ref_base
-            else:
+                    current_type = "SNV"
+
+            if current_type == "Insertion":
+                curr_len = indel_length if indel_length > 0 else random.randint(1, 4)
+                ins_bases = "".join(random.choices(["A", "C", "G", "T"], k=curr_len))
+                ref_seq, alt_seq = ref_base, ref_base + ins_bases
+            elif current_type == "Deletion":
+                curr_len = indel_length if indel_length > 0 else random.randint(1, 4)
+                try:
+                    ref_seq = fasta.fetch(chrom, pos, pos + curr_len + 1).upper()
+                except Exception:
+                    ref_seq = ref_base
+                alt_seq = ref_base
+            else:  # SNV
                 alt_bases = [b for b in ["A", "C", "G", "T"] if b != ref_base]
                 alt_seq = random.choice(alt_bases) if alt_bases else "A"
                 ref_seq = ref_base

src/streamlit_app.py

@@ -254,9 +254,18 @@ uploaded_bed = st.session_state.get("uploaded_bed_file")
 with st.sidebar:
     st.header("Pipeline Parameters")
 
+    enable_batch = st.checkbox(
+        "Enable Batch Controls Mode",
+        value=False,
+        help="Enable this to generate multiple in-silico controls (differing by variant type/VAF) at the exact same coordinates in a single run."
+    )
+
     st.subheader("Sequencing Parameters")
     depth = st.number_input("Target read depth per variant", min_value=1, max_value=10000, value=100, step=10)
-    vaf = st.slider("Variant allele frequency (VAF)", min_value=0.01, max_value=1.0, value=0.20, step=0.01, format="%.2f")
+    if not enable_batch:
+        vaf = st.slider("Variant allele frequency (VAF)", min_value=0.01, max_value=1.0, value=0.20, step=0.01, format="%.2f")
+    else:
+        vaf = 0.20
     read_length = st.number_input("Read length (bp)", min_value=50, max_value=300, value=150, step=10)
 
     st.subheader("Sequencing Technology")
@@ -279,9 +288,28 @@ with st.sidebar:
     indel_interval = st.number_input(
         "Indel interval (0 = SNVs only)",
         min_value=0, max_value=100, value=10, step=1,
-        help="Make every Nth variant an indel. Set to 0 to generate only SNVs.",
+        help="Make every Nth variant an indel. Set to 0 to generate only SNVs. Only active when using Mixed (Random) variants.",
     )
 
+    st.subheader("Variant Specification")
+    if not enable_batch:
+        var_type = st.selectbox(
+            "Variant type",
+            options=["SNV", "Insertion", "Deletion", "Mixed (Random)"],
+            index=3,
+            help="Choose the type of variants to generate. Mixed (Random) alternates between SNVs and Indels."
+        )
+        if var_type in ["Insertion", "Deletion"]:
+            var_length = st.number_input(
+                "Indel length (bp)",
+                min_value=1, max_value=100, value=3, step=1,
+                help="The exact size of the insertion or deletion in base pairs."
+            )
+        else:
+            var_length = 0
+    else:
+        st.info("ℹ️ Batch mode enabled. Variant specifications are configured in the batch table in the main panel.")
+
     st.divider()
     st.subheader("Reference Genome")
     genome_version = st.selectbox(
@@ -479,6 +507,32 @@ if uploaded_bed:
         
         total_selected = len(edited_df[edited_df["Select"] == True])
         st.info(f"Selected {total_selected:,} of {len(df):,} probes ({total_selected/len(df)*100:.1f}%) for variant generation.")
+if uploaded_bed and enable_batch:
+    st.header("1.8 · Configure Batch Runs")
+    st.caption("Define the list of controls to generate. All runs will share the same variant coordinates.")
+    
+    import pandas as pd
+    default_batch = pd.DataFrame([
+        {"Control Name": "control_snv", "VAF": 0.20, "Variant Type": "SNV", "Indel Length": 0},
+        {"Control Name": "control_ins_3bp", "VAF": 0.20, "Variant Type": "Insertion", "Indel Length": 3},
+        {"Control Name": "control_del_5bp", "VAF": 0.20, "Variant Type": "Deletion", "Indel Length": 5},
+    ])
+    
+    if "batch_df" not in st.session_state:
+        st.session_state["batch_df"] = default_batch
+        
+    batch_df = st.data_editor(
+        st.session_state["batch_df"],
+        num_rows="dynamic",
+        use_container_width=True,
+        column_config={
+            "Control Name": st.column_config.TextColumn("Control Name", required=True),
+            "VAF": st.column_config.NumberColumn("VAF", min_value=0.01, max_value=1.0, step=0.01, format="%.2f", required=True),
+            "Variant Type": st.column_config.SelectboxColumn("Variant Type", options=["SNV", "Insertion", "Deletion", "Mixed (Random)"], required=True),
+            "Indel Length": st.column_config.NumberColumn("Indel Length (bp)", min_value=0, max_value=100, step=1, default=0),
+        }
+    )
+    st.session_state["batch_df"] = batch_df
 
 st.divider()
 
@@ -627,91 +681,165 @@ if run_btn:
                 log_func=append_log
             )
 
+        # Resolve runs list
+        if enable_batch:
+            runs_to_process = []
+            for _, row in batch_df.iterrows():
+                runs_to_process.append({
+                    "name": row["Control Name"].strip(),
+                    "vaf": float(row["VAF"]),
+                    "type": "Mixed" if row["Variant Type"].startswith("Mixed") else row["Variant Type"],
+                    "length": int(row["Indel Length"]),
+                })
+        else:
+            runs_to_process = [{
+                "name": "Single Control",
+                "vaf": vaf,
+                "type": "Mixed" if var_type.startswith("Mixed") else var_type,
+                "length": var_length,
+            }]
+
+        # Sanitize and deduplicate names
+        import re
+        seen_names = set()
+        sanitized_runs = []
+        for r in runs_to_process:
+            clean_name = re.sub(r'[^a-zA-Z0-9_]', '_', r["name"])
+            if not clean_name:
+                clean_name = "control"
+            original_clean = clean_name
+            counter = 1
+            while clean_name in seen_names:
+                clean_name = f"{original_clean}_{counter}"
+                counter += 1
+            seen_names.add(clean_name)
+            r["name"] = clean_name
+            sanitized_runs.append(r)
+        runs_to_process = sanitized_runs
+
         append_log("\n============================================")
         append_log("           VARIANT SUMMARY                ")
         append_log("============================================")
-        append_log(f"Total SNVs generated for BAM:       {total_snvs}")
+        append_log(f"Total SNV loci generated:           {total_snvs}")
+        append_log(f"Total runs to execute:              {len(runs_to_process)}")
         append_log("============================================")
 
-        update_progress(0.40, "Generating synthetic BAM...")
-        append_log("\n=== Generating Synthetic BAM ===")
-
-        def bam_progress(fraction, label):
-            update_progress(0.40 + fraction * 0.55, label)
-
-        sorted_bam, output_vcf = pl.generate_synthetic_bam(
-            work_dir=work_dir,
-            snvs_bed=snvs_bed,
-            fasta_path=fasta_path,
-            depth=depth,
-            vaf=vaf,
-            rg_id=rg_id,
-            rg_sm=rg_sm,
-            insert_size=insert_size,
-            insert_std=insert_std,
-            indel_interval=indel_interval,
-            read_length=read_length,
-            sequencing_mode="pcr_amplicon" if seq_mode.startswith("PCR Amplicon") else "hybrid_capture",
-            log_func=append_log,
-            progress_func=bam_progress,
-            compress_fastq=compress_fastq,
-        )
-
-        update_progress(1.0, "Done!")
-        append_log("\n✅ Pipeline complete.")
+        runs_results = {}
+        total_runs = len(runs_to_process)
+        work_dir_name = work_dir.name
+        static_dest_cwd = Path("static") / work_dir_name
+        static_dest_script = Path(__file__).parent / "static" / work_dir_name
 
-        bai_path = Path(str(sorted_bam) + ".bai")
-        vcf_path = Path(output_vcf) if not isinstance(output_vcf, Path) else output_vcf
         igv_bed_path = Path(snvs_bed) if not isinstance(snvs_bed, Path) else snvs_bed
         fully_bed_path = Path(fully_bed) if fully_bed and not isinstance(fully_bed, Path) else fully_bed
+        mane_transcripts_bed_path = Path(mane_transcripts_bed) if mane_transcripts_bed and not isinstance(mane_transcripts_bed, Path) else mane_transcripts_bed
 
-        # Copy to static directories for IGV.js visualization (both root and script-relative)
-        work_dir_name = work_dir.name
-        static_dest_cwd = Path("static") / work_dir_name
-        static_dest_script = Path(__file__).parent / "static" / work_dir_name
-        
-        fastq_r1_path = work_dir / ("synthetic_R1.fastq.gz" if compress_fastq else "synthetic_R1.fastq")
-        fastq_r2_path = work_dir / ("synthetic_R2.fastq.gz" if compress_fastq else "synthetic_R2.fastq")
+        for run_idx, run in enumerate(runs_to_process):
+            run_name = run["name"]
+            run_vaf = run["vaf"]
+            run_type = run["type"]
+            run_length = run["length"]
+
+            append_log(f"\n=== Generating Synthetic BAM for {run_name} (VAF={run_vaf}, {run_type}) ===")
+            
+            def bam_progress(fraction, label):
+                update_progress(0.40 + (run_idx + fraction) / total_runs * 0.55, f"[{run_name}] {label}")
+
+            sorted_bam, output_vcf = pl.generate_synthetic_bam(
+                work_dir=work_dir,
+                snvs_bed=snvs_bed,
+                fasta_path=fasta_path,
+                depth=depth,
+                vaf=run_vaf,
+                rg_id=rg_id,
+                rg_sm=rg_sm,
+                insert_size=insert_size,
+                insert_std=insert_std,
+                indel_interval=indel_interval,
+                read_length=read_length,
+                sequencing_mode="pcr_amplicon" if seq_mode.startswith("PCR Amplicon") else "hybrid_capture",
+                log_func=append_log,
+                progress_func=bam_progress,
+                compress_fastq=compress_fastq,
+                variant_type=run_type,
+                indel_length=run_length
+            )
 
-        for dest in [static_dest_cwd, static_dest_script]:
-            dest.mkdir(parents=True, exist_ok=True)
-            shutil.copy(sorted_bam, dest / "synthetic.sorted.bam")
+            # Resolve output paths in work_dir
+            bai_path = Path(str(sorted_bam) + ".bai")
+            vcf_path = Path(output_vcf) if not isinstance(output_vcf, Path) else output_vcf
+            fastq_r1_path = work_dir / ("synthetic_R1.fastq.gz" if compress_fastq else "synthetic_R1.fastq")
+            fastq_r2_path = work_dir / ("synthetic_R2.fastq.gz" if compress_fastq else "synthetic_R2.fastq")
+
+            # Relocate to run subfolder in work_dir
+            run_work_dir = work_dir / run_name
+            run_work_dir.mkdir(parents=True, exist_ok=True)
+
+            bam_reloc = run_work_dir / "synthetic.sorted.bam"
+            bai_reloc = run_work_dir / "synthetic.sorted.bam.bai"
+            vcf_reloc = run_work_dir / "synthetic.vcf"
+            fastq_r1_reloc = run_work_dir / fastq_r1_path.name
+            fastq_r2_reloc = run_work_dir / fastq_r2_path.name
+
+            shutil.move(str(sorted_bam), str(bam_reloc))
             if bai_path.exists():
-                shutil.copy(bai_path, dest / "synthetic.sorted.bam.bai")
-            shutil.copy(vcf_path, dest / "synthetic.vcf")
-            shutil.copy(igv_bed_path, dest / "igv_variant_navigator.bed")
-            if fully_bed_path and fully_bed_path.exists():
-                shutil.copy(fully_bed_path, dest / "fully_covered_exons.bed")
-            if mane_transcripts_bed and mane_transcripts_bed.exists():
-                shutil.copy(mane_transcripts_bed, dest / "mane_transcripts.bed")
+                shutil.move(str(bai_path), str(bai_reloc))
+            shutil.move(str(vcf_path), str(vcf_reloc))
             if fastq_r1_path.exists():
-                shutil.copy(fastq_r1_path, dest / fastq_r1_path.name)
+                shutil.move(str(fastq_r1_path), str(fastq_r1_reloc))
             if fastq_r2_path.exists():
-                shutil.copy(fastq_r2_path, dest / fastq_r2_path.name)
+                shutil.move(str(fastq_r2_path), str(fastq_r2_reloc))
+
+            # Copy to static subfolder dest / run_name
+            for dest_root in [static_dest_cwd, static_dest_script]:
+                dest = dest_root / run_name
+                dest.mkdir(parents=True, exist_ok=True)
+
+                shutil.copy(bam_reloc, dest / "synthetic.sorted.bam")
+                if bai_reloc.exists():
+                    shutil.copy(bai_reloc, dest / "synthetic.sorted.bam.bai")
+                shutil.copy(vcf_reloc, dest / "synthetic.vcf")
+                shutil.copy(igv_bed_path, dest / "igv_variant_navigator.bed")
+                if fully_bed_path and fully_bed_path.exists():
+                    shutil.copy(fully_bed_path, dest / "fully_covered_exons.bed")
+                if mane_transcripts_bed_path and mane_transcripts_bed_path.exists():
+                    shutil.copy(mane_transcripts_bed_path, dest / "mane_transcripts.bed")
+                if fastq_r1_reloc.exists():
+                    shutil.copy(fastq_r1_reloc, dest / fastq_r1_reloc.name)
+                if fastq_r2_reloc.exists():
+                    shutil.copy(fastq_r2_reloc, dest / fastq_r2_reloc.name)
+
+            runs_results[run_name] = {
+                "stats": stats,
+                "total_snvs": total_snvs,
+                "bam_path": str(bam_reloc),
+                "bai_path": str(bai_reloc) if bai_reloc.exists() else None,
+                "vcf_path": str(vcf_reloc),
+                "igv_bed_path": str(igv_bed_path),
+                "fully_covered_bed_path": str(fully_bed_path) if fully_bed_path else None,
+                "mane_transcripts_bed_path": str(mane_transcripts_bed_path) if mane_transcripts_bed_path else None,
+                "fastq_r1_path": str(fastq_r1_reloc) if fastq_r1_reloc.exists() else None,
+                "fastq_r2_path": str(fastq_r2_reloc) if fastq_r2_reloc.exists() else None,
+                "bam_url": f"/app/static/{work_dir_name}/{run_name}/synthetic.sorted.bam",
+                "bai_url": f"/app/static/{work_dir_name}/{run_name}/synthetic.sorted.bam.bai" if bai_reloc.exists() else None,
+                "vcf_url": f"/app/static/{work_dir_name}/{run_name}/synthetic.vcf",
+                "igv_bed_url": f"/app/static/{work_dir_name}/{run_name}/igv_variant_navigator.bed",
+                "fully_covered_bed_url": f"/app/static/{work_dir_name}/{run_name}/fully_covered_exons.bed" if fully_bed_path and fully_bed_path.exists() else None,
+                "mane_transcripts_bed_url": f"/app/static/{work_dir_name}/{run_name}/mane_transcripts.bed" if mane_transcripts_bed_path and mane_transcripts_bed_path.exists() else None,
+                "fastq_r1_url": f"/app/static/{work_dir_name}/{run_name}/{fastq_r1_reloc.name}" if fastq_r1_reloc.exists() else None,
+                "fastq_r2_url": f"/app/static/{work_dir_name}/{run_name}/{fastq_r2_reloc.name}" if fastq_r2_reloc.exists() else None,
+                "work_dir_name": f"{work_dir_name}/{run_name}",
+                "genome_version": genome_version,
+                "compress_fastq": compress_fastq,
+            }
+
+        update_progress(1.0, "Done!")
+        append_log("\n✅ Pipeline complete.")
 
         # Store URLs and path strings — never load large files into session_state memory
         st.session_state["results"] = {
-            "stats": stats,
-            "total_snvs": total_snvs,
-            "bam_path": str(sorted_bam),
-            "bai_path": str(bai_path) if bai_path.exists() else None,
-            "vcf_path": str(vcf_path),
-            "igv_bed_path": str(igv_bed_path),
-            "fully_covered_bed_path": str(fully_bed_path) if fully_bed_path else None,
-            "mane_transcripts_bed_path": str(mane_transcripts_bed) if mane_transcripts_bed else None,
-            "fastq_r1_path": str(fastq_r1_path) if fastq_r1_path.exists() else None,
-            "fastq_r2_path": str(fastq_r2_path) if fastq_r2_path.exists() else None,
-            "bam_url": f"/app/static/{work_dir_name}/synthetic.sorted.bam",
-            "bai_url": f"/app/static/{work_dir_name}/synthetic.sorted.bam.bai" if bai_path.exists() else None,
-            "vcf_url": f"/app/static/{work_dir_name}/synthetic.vcf",
-            "igv_bed_url": f"/app/static/{work_dir_name}/igv_variant_navigator.bed",
-            "fully_covered_bed_url": f"/app/static/{work_dir_name}/fully_covered_exons.bed" if fully_bed_path and fully_bed_path.exists() else None,
-            "mane_transcripts_bed_url": f"/app/static/{work_dir_name}/mane_transcripts.bed" if mane_transcripts_bed and mane_transcripts_bed.exists() else None,
-            "fastq_r1_url": f"/app/static/{work_dir_name}/{fastq_r1_path.name}" if fastq_r1_path.exists() else None,
-            "fastq_r2_url": f"/app/static/{work_dir_name}/{fastq_r2_path.name}" if fastq_r2_path.exists() else None,
-            "work_dir_name": work_dir_name,
-            "genome_version": genome_version,
-            "compress_fastq": compress_fastq,
+            "is_batch": enable_batch,
+            "runs": runs_results
         }
         st.session_state["log_lines"] = log_lines[:]
 
@@ -723,10 +851,8 @@ if run_btn:
 # ── Results section (persists across reruns via session_state) ────────────────
 
 if "results" in st.session_state:
-    res = st.session_state["results"]
-    stats = res["stats"]
-    total_snvs = res["total_snvs"]
-
+    raw_results = st.session_state["results"]
+    
     st.success("Pipeline completed successfully!")
 
     # Show log if available and pipeline didn't just run
@@ -735,6 +861,25 @@ if "results" in st.session_state:
             st.code("\n".join(st.session_state["log_lines"][-80:]), language=None)
 
     st.header("3 · Results")
+    
+    is_batch = raw_results.get("is_batch", False)
+    if is_batch:
+        run_names = list(raw_results["runs"].keys())
+        selected_run = st.selectbox(
+            "Select control run to view/download",
+            options=run_names,
+            help="Choose which batch control configuration to display and download outputs for."
+        )
+        res = raw_results["runs"][selected_run]
+        
+        # Display run details
+        st.markdown(f"**Selected Run Parameters:** Variant Type: `{res.get('variant_type')}`, VAF: `{res.get('vaf')}`" + 
+                    (f", Indel Length: `{res.get('indel_length')} bp`" if res.get("variant_type") in ["Insertion", "Deletion"] else ""))
+    else:
+        res = list(raw_results["runs"].values())[0]
+
+    stats = res["stats"]
+    total_snvs = res["total_snvs"]
     m1, m2, m3, m4 = st.columns(4)
     m1.metric("Fully Covered Exons", f"{stats['fully_covered']:,}")
     m2.metric("Partially Covered Exons", f"{stats['partially_covered']:,}")

streamlit_app.py

@@ -254,9 +254,18 @@ uploaded_bed = st.session_state.get("uploaded_bed_file")
 with st.sidebar:
     st.header("Pipeline Parameters")
 
+    enable_batch = st.checkbox(
+        "Enable Batch Controls Mode",
+        value=False,
+        help="Enable this to generate multiple in-silico controls (differing by variant type/VAF) at the exact same coordinates in a single run."
+    )
+
     st.subheader("Sequencing Parameters")
     depth = st.number_input("Target read depth per variant", min_value=1, max_value=10000, value=100, step=10)
-    vaf = st.slider("Variant allele frequency (VAF)", min_value=0.01, max_value=1.0, value=0.20, step=0.01, format="%.2f")
+    if not enable_batch:
+        vaf = st.slider("Variant allele frequency (VAF)", min_value=0.01, max_value=1.0, value=0.20, step=0.01, format="%.2f")
+    else:
+        vaf = 0.20
     read_length = st.number_input("Read length (bp)", min_value=50, max_value=300, value=150, step=10)
 
     st.subheader("Sequencing Technology")
@@ -279,9 +288,28 @@ with st.sidebar:
     indel_interval = st.number_input(
         "Indel interval (0 = SNVs only)",
         min_value=0, max_value=100, value=10, step=1,
-        help="Make every Nth variant an indel. Set to 0 to generate only SNVs.",
+        help="Make every Nth variant an indel. Set to 0 to generate only SNVs. Only active when using Mixed (Random) variants.",
     )
 
+    st.subheader("Variant Specification")
+    if not enable_batch:
+        var_type = st.selectbox(
+            "Variant type",
+            options=["SNV", "Insertion", "Deletion", "Mixed (Random)"],
+            index=3,
+            help="Choose the type of variants to generate. Mixed (Random) alternates between SNVs and Indels."
+        )
+        if var_type in ["Insertion", "Deletion"]:
+            var_length = st.number_input(
+                "Indel length (bp)",
+                min_value=1, max_value=100, value=3, step=1,
+                help="The exact size of the insertion or deletion in base pairs."
+            )
+        else:
+            var_length = 0
+    else:
+        st.info("ℹ️ Batch mode enabled. Variant specifications are configured in the batch table in the main panel.")
+
     st.divider()
     st.subheader("Reference Genome")
     genome_version = st.selectbox(
@@ -479,6 +507,32 @@ if uploaded_bed:
         
         total_selected = len(edited_df[edited_df["Select"] == True])
         st.info(f"Selected {total_selected:,} of {len(df):,} probes ({total_selected/len(df)*100:.1f}%) for variant generation.")
+if uploaded_bed and enable_batch:
+    st.header("1.8 · Configure Batch Runs")
+    st.caption("Define the list of controls to generate. All runs will share the same variant coordinates.")
+    
+    import pandas as pd
+    default_batch = pd.DataFrame([
+        {"Control Name": "control_snv", "VAF": 0.20, "Variant Type": "SNV", "Indel Length": 0},
+        {"Control Name": "control_ins_3bp", "VAF": 0.20, "Variant Type": "Insertion", "Indel Length": 3},
+        {"Control Name": "control_del_5bp", "VAF": 0.20, "Variant Type": "Deletion", "Indel Length": 5},
+    ])
+    
+    if "batch_df" not in st.session_state:
+        st.session_state["batch_df"] = default_batch
+        
+    batch_df = st.data_editor(
+        st.session_state["batch_df"],
+        num_rows="dynamic",
+        use_container_width=True,
+        column_config={
+            "Control Name": st.column_config.TextColumn("Control Name", required=True),
+            "VAF": st.column_config.NumberColumn("VAF", min_value=0.01, max_value=1.0, step=0.01, format="%.2f", required=True),
+            "Variant Type": st.column_config.SelectboxColumn("Variant Type", options=["SNV", "Insertion", "Deletion", "Mixed (Random)"], required=True),
+            "Indel Length": st.column_config.NumberColumn("Indel Length (bp)", min_value=0, max_value=100, step=1, default=0),
+        }
+    )
+    st.session_state["batch_df"] = batch_df
 
 st.divider()
 
@@ -627,91 +681,165 @@ if run_btn:
                 log_func=append_log
             )
 
+        # Resolve runs list
+        if enable_batch:
+            runs_to_process = []
+            for _, row in batch_df.iterrows():
+                runs_to_process.append({
+                    "name": row["Control Name"].strip(),
+                    "vaf": float(row["VAF"]),
+                    "type": "Mixed" if row["Variant Type"].startswith("Mixed") else row["Variant Type"],
+                    "length": int(row["Indel Length"]),
+                })
+        else:
+            runs_to_process = [{
+                "name": "Single Control",
+                "vaf": vaf,
+                "type": "Mixed" if var_type.startswith("Mixed") else var_type,
+                "length": var_length,
+            }]
+
+        # Sanitize and deduplicate names
+        import re
+        seen_names = set()
+        sanitized_runs = []
+        for r in runs_to_process:
+            clean_name = re.sub(r'[^a-zA-Z0-9_]', '_', r["name"])
+            if not clean_name:
+                clean_name = "control"
+            original_clean = clean_name
+            counter = 1
+            while clean_name in seen_names:
+                clean_name = f"{original_clean}_{counter}"
+                counter += 1
+            seen_names.add(clean_name)
+            r["name"] = clean_name
+            sanitized_runs.append(r)
+        runs_to_process = sanitized_runs
+
         append_log("\n============================================")
         append_log("           VARIANT SUMMARY                ")
         append_log("============================================")
-        append_log(f"Total SNVs generated for BAM:       {total_snvs}")
+        append_log(f"Total SNV loci generated:           {total_snvs}")
+        append_log(f"Total runs to execute:              {len(runs_to_process)}")
         append_log("============================================")
 
-        update_progress(0.40, "Generating synthetic BAM...")
-        append_log("\n=== Generating Synthetic BAM ===")
-
-        def bam_progress(fraction, label):
-            update_progress(0.40 + fraction * 0.55, label)
-
-        sorted_bam, output_vcf = pl.generate_synthetic_bam(
-            work_dir=work_dir,
-            snvs_bed=snvs_bed,
-            fasta_path=fasta_path,
-            depth=depth,
-            vaf=vaf,
-            rg_id=rg_id,
-            rg_sm=rg_sm,
-            insert_size=insert_size,
-            insert_std=insert_std,
-            indel_interval=indel_interval,
-            read_length=read_length,
-            sequencing_mode="pcr_amplicon" if seq_mode.startswith("PCR Amplicon") else "hybrid_capture",
-            log_func=append_log,
-            progress_func=bam_progress,
-            compress_fastq=compress_fastq,
-        )
-
-        update_progress(1.0, "Done!")
-        append_log("\n✅ Pipeline complete.")
+        runs_results = {}
+        total_runs = len(runs_to_process)
+        work_dir_name = work_dir.name
+        static_dest_cwd = Path("static") / work_dir_name
+        static_dest_script = Path(__file__).parent / "static" / work_dir_name
 
-        bai_path = Path(str(sorted_bam) + ".bai")
-        vcf_path = Path(output_vcf) if not isinstance(output_vcf, Path) else output_vcf
         igv_bed_path = Path(snvs_bed) if not isinstance(snvs_bed, Path) else snvs_bed
         fully_bed_path = Path(fully_bed) if fully_bed and not isinstance(fully_bed, Path) else fully_bed
+        mane_transcripts_bed_path = Path(mane_transcripts_bed) if mane_transcripts_bed and not isinstance(mane_transcripts_bed, Path) else mane_transcripts_bed
 
-        # Copy to static directories for IGV.js visualization (both root and script-relative)
-        work_dir_name = work_dir.name
-        static_dest_cwd = Path("static") / work_dir_name
-        static_dest_script = Path(__file__).parent / "static" / work_dir_name
-        
-        fastq_r1_path = work_dir / ("synthetic_R1.fastq.gz" if compress_fastq else "synthetic_R1.fastq")
-        fastq_r2_path = work_dir / ("synthetic_R2.fastq.gz" if compress_fastq else "synthetic_R2.fastq")
+        for run_idx, run in enumerate(runs_to_process):
+            run_name = run["name"]
+            run_vaf = run["vaf"]
+            run_type = run["type"]
+            run_length = run["length"]
+
+            append_log(f"\n=== Generating Synthetic BAM for {run_name} (VAF={run_vaf}, {run_type}) ===")
+            
+            def bam_progress(fraction, label):
+                update_progress(0.40 + (run_idx + fraction) / total_runs * 0.55, f"[{run_name}] {label}")
+
+            sorted_bam, output_vcf = pl.generate_synthetic_bam(
+                work_dir=work_dir,
+                snvs_bed=snvs_bed,
+                fasta_path=fasta_path,
+                depth=depth,
+                vaf=run_vaf,
+                rg_id=rg_id,
+                rg_sm=rg_sm,
+                insert_size=insert_size,
+                insert_std=insert_std,
+                indel_interval=indel_interval,
+                read_length=read_length,
+                sequencing_mode="pcr_amplicon" if seq_mode.startswith("PCR Amplicon") else "hybrid_capture",
+                log_func=append_log,
+                progress_func=bam_progress,
+                compress_fastq=compress_fastq,
+                variant_type=run_type,
+                indel_length=run_length
+            )
 
-        for dest in [static_dest_cwd, static_dest_script]:
-            dest.mkdir(parents=True, exist_ok=True)
-            shutil.copy(sorted_bam, dest / "synthetic.sorted.bam")
+            # Resolve output paths in work_dir
+            bai_path = Path(str(sorted_bam) + ".bai")
+            vcf_path = Path(output_vcf) if not isinstance(output_vcf, Path) else output_vcf
+            fastq_r1_path = work_dir / ("synthetic_R1.fastq.gz" if compress_fastq else "synthetic_R1.fastq")
+            fastq_r2_path = work_dir / ("synthetic_R2.fastq.gz" if compress_fastq else "synthetic_R2.fastq")
+
+            # Relocate to run subfolder in work_dir
+            run_work_dir = work_dir / run_name
+            run_work_dir.mkdir(parents=True, exist_ok=True)
+
+            bam_reloc = run_work_dir / "synthetic.sorted.bam"
+            bai_reloc = run_work_dir / "synthetic.sorted.bam.bai"
+            vcf_reloc = run_work_dir / "synthetic.vcf"
+            fastq_r1_reloc = run_work_dir / fastq_r1_path.name
+            fastq_r2_reloc = run_work_dir / fastq_r2_path.name
+
+            shutil.move(str(sorted_bam), str(bam_reloc))
             if bai_path.exists():
-                shutil.copy(bai_path, dest / "synthetic.sorted.bam.bai")
-            shutil.copy(vcf_path, dest / "synthetic.vcf")
-            shutil.copy(igv_bed_path, dest / "igv_variant_navigator.bed")
-            if fully_bed_path and fully_bed_path.exists():
-                shutil.copy(fully_bed_path, dest / "fully_covered_exons.bed")
-            if mane_transcripts_bed and mane_transcripts_bed.exists():
-                shutil.copy(mane_transcripts_bed, dest / "mane_transcripts.bed")
+                shutil.move(str(bai_path), str(bai_reloc))
+            shutil.move(str(vcf_path), str(vcf_reloc))
             if fastq_r1_path.exists():
-                shutil.copy(fastq_r1_path, dest / fastq_r1_path.name)
+                shutil.move(str(fastq_r1_path), str(fastq_r1_reloc))
             if fastq_r2_path.exists():
-                shutil.copy(fastq_r2_path, dest / fastq_r2_path.name)
+                shutil.move(str(fastq_r2_path), str(fastq_r2_reloc))
+
+            # Copy to static subfolder dest / run_name
+            for dest_root in [static_dest_cwd, static_dest_script]:
+                dest = dest_root / run_name
+                dest.mkdir(parents=True, exist_ok=True)
+
+                shutil.copy(bam_reloc, dest / "synthetic.sorted.bam")
+                if bai_reloc.exists():
+                    shutil.copy(bai_reloc, dest / "synthetic.sorted.bam.bai")
+                shutil.copy(vcf_reloc, dest / "synthetic.vcf")
+                shutil.copy(igv_bed_path, dest / "igv_variant_navigator.bed")
+                if fully_bed_path and fully_bed_path.exists():
+                    shutil.copy(fully_bed_path, dest / "fully_covered_exons.bed")
+                if mane_transcripts_bed_path and mane_transcripts_bed_path.exists():
+                    shutil.copy(mane_transcripts_bed_path, dest / "mane_transcripts.bed")
+                if fastq_r1_reloc.exists():
+                    shutil.copy(fastq_r1_reloc, dest / fastq_r1_reloc.name)
+                if fastq_r2_reloc.exists():
+                    shutil.copy(fastq_r2_reloc, dest / fastq_r2_reloc.name)
+
+            runs_results[run_name] = {
+                "stats": stats,
+                "total_snvs": total_snvs,
+                "bam_path": str(bam_reloc),
+                "bai_path": str(bai_reloc) if bai_reloc.exists() else None,
+                "vcf_path": str(vcf_reloc),
+                "igv_bed_path": str(igv_bed_path),
+                "fully_covered_bed_path": str(fully_bed_path) if fully_bed_path else None,
+                "mane_transcripts_bed_path": str(mane_transcripts_bed_path) if mane_transcripts_bed_path else None,
+                "fastq_r1_path": str(fastq_r1_reloc) if fastq_r1_reloc.exists() else None,
+                "fastq_r2_path": str(fastq_r2_reloc) if fastq_r2_reloc.exists() else None,
+                "bam_url": f"/app/static/{work_dir_name}/{run_name}/synthetic.sorted.bam",
+                "bai_url": f"/app/static/{work_dir_name}/{run_name}/synthetic.sorted.bam.bai" if bai_reloc.exists() else None,
+                "vcf_url": f"/app/static/{work_dir_name}/{run_name}/synthetic.vcf",
+                "igv_bed_url": f"/app/static/{work_dir_name}/{run_name}/igv_variant_navigator.bed",
+                "fully_covered_bed_url": f"/app/static/{work_dir_name}/{run_name}/fully_covered_exons.bed" if fully_bed_path and fully_bed_path.exists() else None,
+                "mane_transcripts_bed_url": f"/app/static/{work_dir_name}/{run_name}/mane_transcripts.bed" if mane_transcripts_bed_path and mane_transcripts_bed_path.exists() else None,
+                "fastq_r1_url": f"/app/static/{work_dir_name}/{run_name}/{fastq_r1_reloc.name}" if fastq_r1_reloc.exists() else None,
+                "fastq_r2_url": f"/app/static/{work_dir_name}/{run_name}/{fastq_r2_reloc.name}" if fastq_r2_reloc.exists() else None,
+                "work_dir_name": f"{work_dir_name}/{run_name}",
+                "genome_version": genome_version,
+                "compress_fastq": compress_fastq,
+            }
+
+        update_progress(1.0, "Done!")
+        append_log("\n✅ Pipeline complete.")
 
         # Store URLs and path strings — never load large files into session_state memory
         st.session_state["results"] = {
-            "stats": stats,
-            "total_snvs": total_snvs,
-            "bam_path": str(sorted_bam),
-            "bai_path": str(bai_path) if bai_path.exists() else None,
-            "vcf_path": str(vcf_path),
-            "igv_bed_path": str(igv_bed_path),
-            "fully_covered_bed_path": str(fully_bed_path) if fully_bed_path else None,
-            "mane_transcripts_bed_path": str(mane_transcripts_bed) if mane_transcripts_bed else None,
-            "fastq_r1_path": str(fastq_r1_path) if fastq_r1_path.exists() else None,
-            "fastq_r2_path": str(fastq_r2_path) if fastq_r2_path.exists() else None,
-            "bam_url": f"/app/static/{work_dir_name}/synthetic.sorted.bam",
-            "bai_url": f"/app/static/{work_dir_name}/synthetic.sorted.bam.bai" if bai_path.exists() else None,
-            "vcf_url": f"/app/static/{work_dir_name}/synthetic.vcf",
-            "igv_bed_url": f"/app/static/{work_dir_name}/igv_variant_navigator.bed",
-            "fully_covered_bed_url": f"/app/static/{work_dir_name}/fully_covered_exons.bed" if fully_bed_path and fully_bed_path.exists() else None,
-            "mane_transcripts_bed_url": f"/app/static/{work_dir_name}/mane_transcripts.bed" if mane_transcripts_bed and mane_transcripts_bed.exists() else None,
-            "fastq_r1_url": f"/app/static/{work_dir_name}/{fastq_r1_path.name}" if fastq_r1_path.exists() else None,
-            "fastq_r2_url": f"/app/static/{work_dir_name}/{fastq_r2_path.name}" if fastq_r2_path.exists() else None,
-            "work_dir_name": work_dir_name,
-            "genome_version": genome_version,
-            "compress_fastq": compress_fastq,
+            "is_batch": enable_batch,
+            "runs": runs_results
         }
         st.session_state["log_lines"] = log_lines[:]
 
@@ -723,10 +851,8 @@ if run_btn:
 # ── Results section (persists across reruns via session_state) ────────────────
 
 if "results" in st.session_state:
-    res = st.session_state["results"]
-    stats = res["stats"]
-    total_snvs = res["total_snvs"]
-
+    raw_results = st.session_state["results"]
+    
     st.success("Pipeline completed successfully!")
 
     # Show log if available and pipeline didn't just run
@@ -735,6 +861,25 @@ if "results" in st.session_state:
             st.code("\n".join(st.session_state["log_lines"][-80:]), language=None)
 
     st.header("3 · Results")
+    
+    is_batch = raw_results.get("is_batch", False)
+    if is_batch:
+        run_names = list(raw_results["runs"].keys())
+        selected_run = st.selectbox(
+            "Select control run to view/download",
+            options=run_names,
+            help="Choose which batch control configuration to display and download outputs for."
+        )
+        res = raw_results["runs"][selected_run]
+        
+        # Display run details
+        st.markdown(f"**Selected Run Parameters:** Variant Type: `{res.get('variant_type')}`, VAF: `{res.get('vaf')}`" + 
+                    (f", Indel Length: `{res.get('indel_length')} bp`" if res.get("variant_type") in ["Insertion", "Deletion"] else ""))
+    else:
+        res = list(raw_results["runs"].values())[0]
+
+    stats = res["stats"]
+    total_snvs = res["total_snvs"]
     m1, m2, m3, m4 = st.columns(4)
     m1.metric("Fully Covered Exons", f"{stats['fully_covered']:,}")
     m2.metric("Partially Covered Exons", f"{stats['partially_covered']:,}")