diff --git a/workflow/envs/binspreader.yaml b/workflow/envs/binspreader.yaml
new file mode 100644
index 00000000..8b0402ad
--- /dev/null
+++ b/workflow/envs/binspreader.yaml
@@ -0,0 +1,6 @@
+channels:
+  - conda-forge
+  - bioconda
+  - defaults
+dependencies:
+  - binspreader = 3.16
\ No newline at end of file
diff --git a/workflow/rules/binspreader.smk b/workflow/rules/binspreader.smk
new file mode 100644
index 00000000..42217a1e
--- /dev/null
+++ b/workflow/rules/binspreader.smk
@@ -0,0 +1,104 @@
+
+rule rename_paths:
+    input:
+        paths="{{sample}}/assembly/{sequences}.paths".format(
+            sequences="scaffolds" if config["spades_use_scaffolds"] else "contigs"
+            ),
+        filtered_contigs = get_assembly
+    output:
+        "{sample}/assembly/renamed_assembly.paths"
+    run:
+
+        with open(output[0],"w") as out_handle, open(input.paths) as in_handle:
+            for line in in_handle:
+                if line.startswith("NODE"):
+
+                    # expect"NODE_1_length_196862_cov_1048.163282"
+                    _,nr,_ = line.split("_",2)
+
+                    out_handle.write(f"{wildcards.sample}_{nr}\n")
+                else:
+                    out_handle.write(line)
+
+
+
+
+
+rule prepare_yaml_for_binspreader:
+    input:
+        reads=get_quality_controlled_reads
+    output:
+        temp("Intermediate/binspreader/dataset_yamls/{sample}.yaml")
+    run:
+        assert len(input.reads)>=2
+
+        import yaml
+        from os.path import abspath
+
+        data = [
+            {
+                'orientation': 'fr',
+                'type': 'paired-end',
+                'right reads': [abspath(input.reads[0])],
+                'left reads': [abspath(input.reads[1])]
+            }
+        ]
+
+        with open(output[0], 'w') as file:
+            yaml.dump(data, file, default_style='"', default_flow_style=False)
+
+            
+
+
+rule binspreader:
+    input:
+        #get_quality_controlled_reads,
+        graph="{sample}/assembly/assembly_graph_after_simplification.gfa",
+        #yaml = ancient(rules.prepare_yaml_for_binspreader.output[0]),
+        paths = "{sample}/assembly/renamed_assembly.paths",
+        bins = "{sample}/binning/{binner_before}/cluster_attribution.tsv",
+    output:
+        directory("Intermediate/binspreader/after_{binner_before}/outputs/{sample}")
+    log:
+        "logs/binspreader/after_{binner_before}/run_binspreader/{sample}.log"
+    conda:
+        "../envs/binspreader.yaml"
+    threads:
+        config["threads"]
+    shell:
+        "bin-refine "
+        " {input.graph} " 
+        " {input.bins} "
+        " {output} " 
+        " --paths {input.paths} "
+        " --no-unbinned-bin "
+        "  -Rcorr "
+        " -m " # allow multiple bin assignment
+        " --tmp-dir {resources.tmpdir}/binspreader_{wildcards.sample}_{wildcards.binner_before} "
+        " &> {log} "
+
+        #"  " # only part of graph is binned
+        #" --reads "
+        #" --dataset {input.yaml} "
+        # " --sparse-propagation"
+
+
+rule parse_binspreader:
+    input:
+        rules.binspreader.output
+    output:
+        "{sample}/binning/binsp{binner_before}/cluster_attribution.tsv"
+    run:
+
+        input_file = Path(input[0])/"binning.tsv"
+
+        with open(output[0],"w") as out_handle, open(input_file) as in_handle:
+            for line in in_handle:
+                # contig tab binname1 tab binname2 ...
+
+                contig,bin_names = line.strip().split('\t',1)
+
+                for bin in bin_names.split():
+                    new_bin_name = bin.replace(wildcards.binner_before,"binsp"+wildcards.binner_before)
+
+                    out_handle.write(f"{contig}\t{new_bin_name}\n")
\ No newline at end of file