diff --git a/backend/src/api_types.py b/backend/src/api_types.py
index 69152eac..9797ee93 100644
--- a/backend/src/api_types.py
+++ b/backend/src/api_types.py
@@ -62,3 +62,8 @@ class EditBody(CamelModel):
     new_species_name: str
     new_content: str | None = None
     new_refs: str | None = None
+
+
+class SimilarProtein(CamelModel):
+    name: str
+    prob: float
diff --git a/backend/src/foldseek.py b/backend/src/foldseek.py
new file mode 100644
index 00000000..cd0a4aad
--- /dev/null
+++ b/backend/src/foldseek.py
@@ -0,0 +1,121 @@
+import subprocess
+import logging as log
+
+EXTERNAL_DATABASES = [
+    "Alphafold/UniProt",
+    "Alphafold/UniProt50",
+    "Alphafold/Proteome",
+    "Alphafold/Swiss",
+    "ESMAtlas30",
+    "PDB",
+]
+
+
+def bash_cmd(cmd: str | list[str]) -> str:
+    return subprocess.check_output(cmd, shell=True).decode()
+
+
+def to_columnar_array(arr: list[list]) -> list[list]:
+    columnar = []
+    for i in range(len(arr[0])):
+        columnar.append([])
+        for j in range(len(arr)):
+            columnar[i].append(arr[j][i])
+    return columnar
+
+
+def parse_output(filepath: str) -> list[list]:
+    with open(filepath, "r") as f:
+        lines = f.readlines()
+
+        parsed_lines = []
+        for line in lines:
+            parsed_line = []
+            for column in line.strip("\n").split("\t"):
+                try:
+                    column = float(column)
+                except ValueError:
+                    pass
+                parsed_line.append(column)
+            parsed_lines.append(parsed_line)
+
+    return parsed_lines
+
+
+def easy_search(
+    query: str,
+    target: str,
+    out_format: list[str] = ["query", "target", "prob"],
+    out_file=".foldseek_cache/output",
+    temp_dir=".foldseek_cache",
+    print_stdout=False,
+    foldseek_executable="./foldseek/bin/foldseek",
+    columnar=False,
+) -> list[list]:
+    """easy_search just calls foldseek easy-search under the hood
+    TODO: use pybind to call the C++ function instead
+
+    Returns:
+        list[list]: a list of the matches from the search
+    """
+
+    # Then call the easy-search
+    flags = f"--format-output {','.join(out_format)}" if len(out_format) > 0 else ""
+    cmd = f"{foldseek_executable} easy-search {query} {target} {out_file} {temp_dir} {flags}"
+    try:
+        stdout = bash_cmd(cmd)
+    except Exception as e:
+        log.warn(e)
+        return []
+
+    if print_stdout:
+        log.warn(stdout)
+
+    if columnar:
+        return to_columnar_array(parse_output(out_file))
+    else:
+        return parse_output(out_file)
+
+
+def create_db(
+    dir: str,
+    db_name: str,
+    foldseek_executable="foldseek",
+    print_stdout=False,
+    temp_dir=".foldseek_cache",
+):
+    # don't continue unless they actually have foldseek installed
+    if bash_cmd(f"which {foldseek_executable}") == "":
+        raise Exception("foldseek not found in PATH")
+
+    # check that our dir exists
+    try:
+        bash_cmd(f"ls {dir}")
+    except Exception:
+        if dir not in EXTERNAL_DATABASES:
+            raise Exception(f"Directory {dir} not found")
+
+    # if database already exists, don't create another
+    try:
+        bash_cmd(f"ls {db_name}")
+    except Exception:
+        if dir not in EXTERNAL_DATABASES:
+            cmd = f"{foldseek_executable} createdb {dir} {db_name}"
+        else:
+            cmd = f"{foldseek_executable} databases {dir} {db_name} {temp_dir}"
+        stdout = bash_cmd(cmd)
+        if print_stdout:
+            print(stdout)
+
+    return db_name
+
+
+if __name__ == "__main__":
+    # search each protein in test_examples/ with every other one
+    test_targets = create_db(dir="test_examples", db_name="test")
+    output = easy_search(
+        query=test_targets,
+        target=test_targets,
+        out_format=["query", "target", "prob"],
+    )
+    print(output)
diff --git a/backend/src/protein.py b/backend/src/protein.py
index 14f6c9f7..74783244 100644
--- a/backend/src/protein.py
+++ b/backend/src/protein.py
@@ -49,6 +49,10 @@ def pdb_file_name(protein_name: str):
     )
 
 
+def revert_pdb_filename(file_name: str):
+    return file_name.replace(".pdb", "").replace("_", " ")
+
+
 def parse_protein_pdb(name: str, file_contents: str = "", encoding="str"):
     if encoding == "str":
         return PDB(file_contents, name)
diff --git a/backend/src/server.py b/backend/src/server.py
index 953c4676..465a577d 100644
--- a/backend/src/server.py
+++ b/backend/src/server.py
@@ -2,16 +2,31 @@
 import os
 from io import BytesIO
 from fastapi.responses import FileResponse, StreamingResponse
-from .api_types import ProteinEntry, UploadBody, UploadError, EditBody
+from .api_types import ProteinEntry, UploadBody, UploadError, EditBody, SimilarProtein
 from .db import Database, bytea_to_str, str_to_bytea
-from .protein import parse_protein_pdb, pdb_file_name, protein_name_found, pdb_to_fasta
+from .protein import (
+    parse_protein_pdb,
+    pdb_file_name,
+    protein_name_found,
+    pdb_to_fasta,
+    revert_pdb_filename,
+)
 from .setup import disable_cors, init_fastapi_app
+from .foldseek import easy_search
 
 
 app = init_fastapi_app()
 disable_cors(app, origins=[os.environ["PUBLIC_FRONTEND_URL"]])
 
 
+@app.get("/similar-venome/{protein_name:str}", response_model=list[SimilarProtein])
+def get_venome_proteins(protein_name: str):
+    query_name = pdb_file_name(protein_name)
+    target_folder = "src/data/pdbAlphaFold/"
+    similar = easy_search(query_name, target_folder, out_format=["target", "prob"])
+    return [SimilarProtein(name=revert_pdb_filename(s[0]), prob=s[1]) for s in similar]
+
+
 @app.get("/pdb/{protein_name:str}")
 def get_pdb_file(protein_name: str):
     if protein_name_found(protein_name):