diff --git a/.github/workflows/build.yml b/.github/workflows/build.yml
new file mode 100644
index 0000000..c7945e3
--- /dev/null
+++ b/.github/workflows/build.yml
@@ -0,0 +1,29 @@
+name: Docker Build Test
+
+on:
+ push:
+ branches:
+ master
+ pull_request:
+ branches:
+ master
+ schedule:
+ - cron: '0 0 * * 1'
+
+jobs:
+ build-test:
+ runs-on: ubuntu-latest
+ strategy:
+ max-parallel: 4
+ matrix:
+ python-version: [3.8]
+# TODO: We probably should switch to using the Docker version.
+ steps:
+ - uses: actions/checkout@v2
+ - name: Set up Python ${{ matrix.python-version }}
+ uses: actions/setup-python@v2
+ with:
+ python-version: ${{ matrix.python-version }}
+ - name: Build Docker
+ run: |
+ docker build .
diff --git a/Dockerfile b/Dockerfile
index fa855a9..24dda63 100644
--- a/Dockerfile
+++ b/Dockerfile
@@ -1,14 +1,15 @@
-FROM continuumio/miniconda3:4.8.2
+FROM continuumio/miniconda3:4.10.3
MAINTAINER Mingxun Wang "mwang87@gmail.com"
WORKDIR /app
RUN apt-get update -y && \
apt-get install -y libxrender-dev && \
apt-get install -y git-core
-RUN conda create -y -n usi -c conda-forge -c bioconda -c defaults celery \
+RUN conda install -c conda-forge mamba
+RUN mamba create -y -n usi -c conda-forge -c bioconda -c defaults celery \
dash=1.20.0 dash-bootstrap-components=0.9.2 flask gunicorn \
joblib matplotlib numba numpy openssl qrcode rdkit requests \
- requests-cache scipy spectrum_utils werkzeug
+ requests-cache scipy spectrum_utils==0.3.5 werkzeug==2.0.0
RUN /bin/bash -c 'source activate usi && pip install "git+https://github.com/berlinguyinca/spectra-hash.git#subdirectory=python" && pip install celery-once'
RUN echo "source activate usi" > ~/.bashrc
diff --git a/Makefile b/Makefile
index 5d5f5bc..586efe7 100644
--- a/Makefile
+++ b/Makefile
@@ -23,19 +23,19 @@ clear-cache:
#Docker Compose
server-compose-interactive:
- docker-compose build
- docker-compose up
+ docker-compose --compatibility build
+ docker-compose --compatibility up
server-compose:
- docker-compose build
- docker-compose up -d
+ docker-compose --compatibility build
+ docker-compose --compatibility up -d
server-compose-production-interactive:
- docker-compose build
+ docker-compose --compatibility build
docker-compose -f docker-compose.yml -f docker-compose-production.yml --compatibility up
server-compose-production:
- docker-compose build
+ docker-compose --compatibility build
docker-compose -f docker-compose.yml -f docker-compose-production.yml --compatibility up -d
attach:
diff --git a/docker-compose-production.yml b/docker-compose-production.yml
index 5423467..827f08a 100644
--- a/docker-compose-production.yml
+++ b/docker-compose-production.yml
@@ -5,9 +5,11 @@ services:
- default
- nginx-net
environment:
- VIRTUAL_HOST: metabolomics-usi.ucsd.edu,metabolomics-usi.gnps2.org
- VIRTUAL_PORT: 5087
- LETSENCRYPT_HOST: metabolomics-usi.ucsd.edu,metabolomics-usi.gnps2.org
+ #VIRTUAL_HOST: metabolomics-usi.ucsd.edu,metabolomics-usi.gnps2.org
+ VIRTUAL_HOST: metabolomics-usi.gnps2.org
+ VIRTUAL_PORT: 5000
+ #LETSENCRYPT_HOST: metabolomics-usi.ucsd.edu,metabolomics-usi.gnps2.org
+ LETSENCRYPT_HOST: metabolomics-usi.gnps2.org
LETSENCRYPT_EMAIL: mwang87@gmail.com
command: /app/run_server.sh
deploy:
diff --git a/metabolomics_spectrum_resolver/dashinterface.py b/metabolomics_spectrum_resolver/dashinterface.py
index 5069ef2..26abbd9 100644
--- a/metabolomics_spectrum_resolver/dashinterface.py
+++ b/metabolomics_spectrum_resolver/dashinterface.py
@@ -426,7 +426,7 @@
dbc.CardHeader(html.H5("Contributors")),
dbc.CardBody(
[
- "Mingxun Wang, PhD – UC San Diego",
+ "Mingxun Wang, PhD – UC Riverside",
html.Br(),
"Wout Bittremieux, PhD – UC San Diego",
html.Br(),
diff --git a/metabolomics_spectrum_resolver/parsing.py b/metabolomics_spectrum_resolver/parsing.py
index f488e4c..c0b572f 100644
--- a/metabolomics_spectrum_resolver/parsing.py
+++ b/metabolomics_spectrum_resolver/parsing.py
@@ -4,6 +4,8 @@
from typing import Tuple
import requests
+import pandas as pd
+from io import StringIO
import urllib.parse
import spectrum_utils.spectrum as sus
import splash
@@ -25,7 +27,7 @@
# PXLnnnnnn
# Unofficial: MASSIVEKB
# https://github.com/HUPO-PSI/usi/blob/master/CollectionIdentifiers.md
- r":(MSV\d{9}|PXD\d{6}|PXL\d{6}|RPXD\d{6}|MassIVE)"
+ r":(MSV\d{9}|PXD\d{6}|PXL\d{6}|RPXD\d{6}|ST\d{6}|MassIVE)"
# msRun identifier
r":(.*)"
# index flag
@@ -116,6 +118,8 @@ def parse_usi(usi: str) -> Tuple[sus.MsmsSpectrum, str, str]:
spectrum, source_link = _parse_ms2lda(usi)
elif collection == "motifdb":
spectrum, source_link = _parse_motifdb(usi)
+ elif collection.startswith("st"):
+ spectrum, source_link = _parse_metabolomics_workbench(usi)
else:
raise UsiError(f"Unknown USI collection: {match.group(1)}", 400)
splash_key = splash_builder.splash(
@@ -596,6 +600,57 @@ def _parse_motifdb(usi: str) -> Tuple[sus.MsmsSpectrum, str]:
raise UsiError("Unknown MOTIFDB USI", 404)
+# Parse GNPS library.
+def _parse_metabolomics_workbench(usi: str) -> Tuple[sus.MsmsSpectrum, str]:
+ match = _match_usi(usi)
+ accession = match.group(1)
+ filename = match.group(2)
+ index_flag = match.group(3)
+ index = match.group(4)
+
+ if index_flag.lower() != "scan":
+ raise UsiError(
+ "Currently supported MW index flags: scan", 400
+ )
+ try:
+ request_url = (
+ f"https://www.metabolomicsworkbench.org/"
+ f"data/ms2.php?A={accession}.zip"
+ f"&F={urllib.parse.quote_plus(filename)}&S={index}"
+ )
+
+ # TODO: Do some extra exception handling if we don't find the filename directly. We might need to his another API to get the full filename
+ # Given the just the basename
+
+ lookup_request = requests.get(request_url, timeout=timeout)
+ lookup_request.raise_for_status()
+
+ response_text = lookup_request.text
+ response_text = (response_text.replace("", "").replace("", "").lstrip().rstrip())
+
+ # Parsing the MW Response
+ precursor_mz = float(response_text.split("\n")[0].split(":")[-1].replace("\"", ""))
+ charge = int(response_text.split("\n")[2].split(":")[-1].replace("\"", ""))
+ peaks_df = pd.read_csv(StringIO(response_text), sep=r" +", skiprows=4)
+ mz = list(peaks_df["m/z"])
+ intensity = list(peaks_df["intensity"])
+
+ source_link = (
+ f"https://www.metabolomicsworkbench.org/"
+ f"data/DRCCMetadata.php?Mode=Study&StudyID={accession}&StudyType=MS&ResultType=1"
+ )
+
+ spectrum = sus.MsmsSpectrum(
+ usi,
+ float(precursor_mz),
+ int(charge),
+ mz,
+ intensity,
+ )
+ return spectrum, source_link
+ except requests.exceptions.HTTPError:
+ raise UsiError("Unknown MW USI", 404)
+
def _parse_sequence(peptide: str, peptide_clean: str) -> Tuple[str, str, list]:
# Parse out gapped sequence (e.g. X+129.04259), faking it
# with Glycine as the base residue and adding more mods to
diff --git a/metabolomics_spectrum_resolver/templates/homepage.html b/metabolomics_spectrum_resolver/templates/homepage.html
index 2e827bd..a6cf1ad 100644
--- a/metabolomics_spectrum_resolver/templates/homepage.html
+++ b/metabolomics_spectrum_resolver/templates/homepage.html
@@ -136,7 +136,7 @@
diff --git a/requirements.txt b/requirements.txt
index fdddf45..62ea414 100644
--- a/requirements.txt
+++ b/requirements.txt
@@ -19,5 +19,5 @@ requests
requests_cache
scipy
spectrum_utils
-werkzeug
-git+https://github.com/berlinguyinca/spectra-hash.git#subdirectory=python
+werkzeug==2.0.0
+git+https://github.com/berlinguyinca/spectra-hash.git#subdirectory=python
\ No newline at end of file
diff --git a/test/usi_test_data.py b/test/usi_test_data.py
index b80eee0..670b4b2 100644
--- a/test/usi_test_data.py
+++ b/test/usi_test_data.py
@@ -29,6 +29,8 @@
"mzspec:MassIVE:TASK-f4b86b150a164ee4a440b661e97a7193-spectra/specs_ms.mgf:scan:287215:HPYFYAPELLF[-10.059]FAKR/3",
# MassIVE Task USIs disguised as GNPS Task USIs
"mzspec:GNPS:TASK-f4b86b150a164ee4a440b661e97a7193-spectra/specs_ms.mgf:scan:287215:HPYFYAPELLF[-10.059]FAKR/3",
+ # Metabolomics Workbench USIs
+ "mzspec:ST000003:StemCell+Data+and+Raw+Files/iPSC-T1R1:scan:3",
# Legacy cases.
"mzspec:GNPSTASK-c95481f0c53d42e78a61bf899e9f9adb:spectra/specs_ms.mgf:scan:1943",
"mzspec:GNPSTASK-64b22841ab3548f987b3cfc18696a581:spectra/specs_ms.mgf:scan:1469",