feat: BTB

.github/workflows/build-containers.yml

@@ -43,6 +43,11 @@ jobs:
     with:
       path: docker-wrappers/DOMINO
       container: reedcompbio/domino
+  build-and-remove-btb:
+    uses: "./.github/workflows/build-and-remove-template.yml"
+    with:
+      path: docker-wrappers/BowTieBuilder
+      container: reedcompbio/bowtiebuilder
   build-and-remove-cytoscape:
     uses: "./.github/workflows/build-and-remove-template.yml"
     with:

config/config.yaml

@@ -110,6 +110,10 @@ algorithms:
         alpha: [0.85]
         threshold: [100, 200]
 
+  - name: "bowtiebuilder"
+    params:
+      include: true
+
 # Here we specify which pathways to run and other file location information.
 # DataLoader.py can currently only load a single dataset
 # Assume that if a dataset label does not change, the lists of associated input files do not change

docker-wrappers/BowTieBuilder/Dockerfile

@@ -0,0 +1,5 @@
+FROM python:3.8-bullseye
+
+WORKDIR /btb
+RUN wget https://raw.githubusercontent.com/Reed-CompBio/BowTieBuilder-Algorithm/dd8519cd8a8397c0e0724106f498b6002d3f7be2/btb.py
+RUN pip install networkx==2.8

docker-wrappers/BowTieBuilder/README.md

@@ -0,0 +1,15 @@
+# BowTieBuilder Docker image
+
+A Docker image for [BowTieBuilder](https://github.com/Reed-CompBio/BowTieBuilder-Algorithm) that is available on [DockerHub](https://hub.docker.com/repository/docker/reedcompbio/bowtiebuilder).
+
+To create the Docker image run:
+```
+docker build -t reedcompbio/bowtiebuilder:v2 -f Dockerfile .
+```
+from this directory.
+
+## Original Paper
+
+The original paper for [BowTieBuilder] can be accessed here:
+
+Supper, J., Spangenberg, L., Planatscher, H. et al. BowTieBuilder: modeling signal transduction pathways. BMC Syst Biol 3, 67 (2009). https://doi.org/10.1186/1752-0509-3-67

spras/btb.py

@@ -0,0 +1,162 @@
+from pathlib import Path
+
+from spras.containers import prepare_volume, run_container_and_log
+from spras.interactome import (
+    convert_undirected_to_directed,
+    reinsert_direction_col_directed,
+)
+from spras.prm import PRM
+from spras.util import (
+    add_rank_column,
+    duplicate_edges,
+    raw_pathway_df,
+)
+
+__all__ = ['BowTieBuilder']
+
+"""
+BTB will construct a BowTie-shaped graph from the provided input file.
+BTB works with directed and undirected graphs.
+It generates a graph connecting multiple source nodes to multiple target nodes with the minimal number of intermediate nodes as possible.
+
+Expected raw edge file format:
+Interactor1     Interactor2     Weight
+"""
+
+class BowTieBuilder(PRM):
+    required_inputs = ['sources', 'targets', 'edges']
+
+    #generate input taken from meo.py beacuse they have same input requirements
+    @staticmethod
+    def generate_inputs(data, filename_map):
+        """
+        Access fields from the dataset and write the required input files
+        @param data: dataset
+        @param filename_map: a dict mapping file types in the required_inputs to the filename for that type
+        @return:
+        """
+        for input_type in BowTieBuilder.required_inputs:
+            if input_type not in filename_map:
+                raise ValueError(f"{input_type} filename is missing")
+
+        # Get sources and write to file, repeat for targets
+        # Does not check whether a node is a source and a target
+        for node_type in ['sources', 'targets']:
+            nodes = data.request_node_columns([node_type])
+            if nodes is None:
+                raise ValueError(f'No {node_type} found in the node files')
+
+            # TODO test whether this selection is needed, what values could the column contain that we would want to
+            # include or exclude?
+            nodes = nodes.loc[nodes[node_type]]
+            if node_type == "sources":
+                nodes.to_csv(filename_map["sources"], sep= '\t', index=False, columns=['NODEID'], header=False)
+            elif node_type == "targets":
+                nodes.to_csv(filename_map["targets"], sep= '\t', index=False, columns=['NODEID'], header=False)
+
+
+        # Create network file
+        edges = data.get_interactome()
+
+        # Format into directed graph (BTB uses the nx.DiGraph constructor internally)
+        edges = convert_undirected_to_directed(edges)
+
+        edges.to_csv(filename_map["edges"], sep="\t", index=False,
+                                      columns=["Interactor1", "Interactor2", "Weight"],
+                                      header=False)
+
+
+
+    # Skips parameter validation step
+    @staticmethod
+    def run(sources=None, targets=None, edges=None, output_file=None, container_framework="docker"):
+        """
+        Run BTB with Docker
+        @param sources:  input source file (required)
+        @param targets:  input target file (required)
+        @param edges:    input edge file (required)
+        @param output_file: path to the output pathway file (required)
+        @param container_framework: choose the container runtime framework, currently supports "docker" or "singularity" (optional)
+        """
+
+        # Tests for pytest (docker container also runs this)
+        # Testing out here avoids the trouble that container errors provide
+
+        if not sources or not targets or not edges or not output_file:
+            raise ValueError('Required BowTieBuilder arguments are missing')
+
+        if not Path(sources).exists() or not Path(targets).exists() or not Path(edges).exists():
+            raise ValueError('Missing input file')
+
+        # Testing for btb index errors
+        # TODO: This error will never actually occur if the inputs are passed through
+        # `generate_inputs`. See the discussion about removing this or making this a habit at
+        # https://github.com/Reed-CompBio/spras/issues/306.
+        with open(edges, 'r') as edge_file:
+            try:
+                for line in edge_file:
+                    line = line.strip().split('\t')[2]
+
+            except Exception as err:
+                # catches a much harder to debug error in BTB.
+                raise IndexError("BTB edges are not formatted correctly") from err
+
+        work_dir = '/btb'
+
+        # Each volume is a tuple (src, dest)
+        volumes = list()
+
+        bind_path, source_file = prepare_volume(sources, work_dir)
+        volumes.append(bind_path)
+
+        bind_path, target_file = prepare_volume(targets, work_dir)
+        volumes.append(bind_path)
+
+        bind_path, edges_file = prepare_volume(edges, work_dir)
+        volumes.append(bind_path)
+
+        # Use its --output argument to set the output file prefix to specify an absolute path and prefix
+        out_dir = Path(output_file).parent
+        out_dir.mkdir(parents=True, exist_ok=True)
+        bind_path, mapped_out_dir = prepare_volume(str(out_dir), work_dir)
+        volumes.append(bind_path)
+        mapped_out_prefix = mapped_out_dir + '/raw-pathway.txt'  # Use posix path inside the container
+
+        command = ['python',
+                   'btb.py',
+                   '--edges',
+                   edges_file,
+                   '--sources',
+                   source_file,
+                   '--targets',
+                   target_file,
+                   '--output_file',
+                   mapped_out_prefix]
+
+        container_suffix = "bowtiebuilder:v2"
+        run_container_and_log('BowTieBuilder',
+                              container_framework,
+                              container_suffix,
+                              command,
+                              volumes,
+                              work_dir)
+        # Output is already written to raw-pathway.txt file
+
+
+    @staticmethod
+    def parse_output(raw_pathway_file, standardized_pathway_file, params):
+        """
+        Convert a predicted pathway into the universal format
+        @param raw_pathway_file: pathway file produced by an algorithm's run function
+        @param standardized_pathway_file: the same pathway written in the universal format
+        """
+        # TODO: consider using multiple raw_pathway_files
+        df = raw_pathway_df(raw_pathway_file, sep='\t', header=0)
+        if not df.empty:
+            df = add_rank_column(df)
+            df = reinsert_direction_col_directed(df)
+            df.columns = ['Node1', 'Node2', 'Rank', 'Direction']
+            df, has_duplicates = duplicate_edges(df)
+            if has_duplicates:
+                print(f"Duplicate edges were removed from {raw_pathway_file}")
+        df.to_csv(standardized_pathway_file, index=False, sep='\t', header=True)

spras/runner.py

@@ -2,6 +2,7 @@
 
 # supported algorithm imports
 from spras.allpairs import AllPairs as allpairs
+from spras.btb import BowTieBuilder as bowtiebuilder
 from spras.dataset import Dataset
 from spras.domino import DOMINO as domino
 from spras.meo import MEO as meo

test/BowTieBuilder/expected/bidirectional-output.txt

@@ -0,0 +1,3 @@
+Node1	Node2
+S1	A
+A	T1

test/BowTieBuilder/expected/btb-output.txt

@@ -0,0 +1,5 @@
+Node1	Node2
+S1	A
+S2	A
+A	T1
+A	T2

test/BowTieBuilder/expected/disjoint-output.txt

@@ -0,0 +1,6 @@
+Node1	Node2
+S1	A
+S2	C
+C	T2
+A	B
+B	T1

test/BowTieBuilder/expected/empty-output.txt

@@ -0,0 +1 @@
+Node1	Node2

test/BowTieBuilder/expected/loop-output.txt

@@ -0,0 +1,6 @@
+Node1	Node2
+S1	A
+A	B
+B	T1
+T1	C
+C	T2

test/BowTieBuilder/expected/source-to-source-disjoint-output.txt

@@ -0,0 +1,6 @@
+Node1	Node2
+S1	A
+S1	S2
+S2	B
+A	T1
+B	T2

test/BowTieBuilder/expected/source-to-source-output.txt

@@ -0,0 +1,4 @@
+Node1	Node2
+S1	A
+A	T1
+A	T2

test/BowTieBuilder/expected/source-to-source2-output.txt

@@ -0,0 +1,5 @@
+Node1	Node2
+S1	A
+S2	S1
+A	T1
+A	T2

test/BowTieBuilder/expected/weighted-output.txt

@@ -0,0 +1,3 @@
+Node1	Node2
+S1	A
+A	T1

test/BowTieBuilder/input/bad-edges.txt

@@ -0,0 +1,6 @@
+A	D	5
+B	D	1.3
+C	0.4
+D	E	4.5
+D	F	2
+D	G	3.2

test/BowTieBuilder/input/bidirectional-edges.txt

@@ -0,0 +1,4 @@
+S1	A	1
+A	T1	1
+A	S1	1
+T1	A	1

test/BowTieBuilder/input/btb-edges.txt

@@ -0,0 +1,5 @@
+S1	A	1
+S1	S2	1
+S2	A	1
+A	T1	1
+A	T2	1

test/BowTieBuilder/input/btb-sources.txt

@@ -0,0 +1,2 @@
+S1
+S2

test/BowTieBuilder/input/btb-targets.txt

@@ -0,0 +1,2 @@
+T1
+T2

test/BowTieBuilder/input/disjoint-edges.txt

@@ -0,0 +1,5 @@
+S1	A	1
+A	B	1
+B	T1	1
+S2	C	1
+C	T2	1

test/BowTieBuilder/input/disjoint-sources.txt

@@ -0,0 +1,3 @@
+S1
+S2
+S3

test/BowTieBuilder/input/disjoint-targets.txt

@@ -0,0 +1,3 @@
+T1
+T2
+T3

test/BowTieBuilder/input/disjoint2-edges.txt

@@ -0,0 +1,6 @@
+S1	A	1
+A	B	1
+B	T1	1
+S2	C	1
+C	T2	1
+S3	D	1

test/BowTieBuilder/input/edges1.txt

@@ -0,0 +1,6 @@
+A	D	5
+B	D	1.3
+C	D	0.4
+D	E	4.5
+D	F	2
+D	G	3.2

test/BowTieBuilder/input/loop-edges.txt

@@ -0,0 +1,6 @@
+S1	A	1
+A	B	1
+B	T1	1
+T1	C	1
+C	T2	1
+T2	S1	1

test/BowTieBuilder/input/source-to-source-disjoint-edges.txt

@@ -0,0 +1,5 @@
+S1	S2	1
+S1	A	1
+A	T1	1
+S2	B	1
+B	T2	1

test/BowTieBuilder/input/source-to-source-edges.txt

@@ -0,0 +1,4 @@
+S1	A	1
+S1	S2	1
+A	T1	1
+A	T2	1

test/BowTieBuilder/input/source-to-source2-edges.txt

@@ -0,0 +1,4 @@
+S1	A	1
+S2	S1	1
+A	T1	1
+A	T2	1

test/BowTieBuilder/input/source1.txt

@@ -0,0 +1,3 @@
+A
+B
+C

test/BowTieBuilder/input/target-to-source-edges.txt

@@ -0,0 +1,2 @@
+A	S1	1
+T1	A	1

test/BowTieBuilder/input/target1.txt

@@ -0,0 +1,3 @@
+E
+F
+G

test/BowTieBuilder/input/weight-one-edges.txt

@@ -0,0 +1,4 @@
+S1	A	1
+A	T1	1
+S1	B	0.5
+B	T1	0.5

test/BowTieBuilder/input/weighted-edges.txt

@@ -0,0 +1,4 @@
+S1	A	0.9
+A	T1	0.9
+S1	B	0.5
+B	T1	0.5