Drop conda dependency: replace hmmer/prodigal/seqkit with pyhmmer/pyrodigal/Python

.bioconda/meta.yaml

@@ -24,14 +24,13 @@ requirements:
     - poetry
     - pip
   run:
-    - hmmer
-    - prodigal
+    - pyhmmer >=0.10
+    - pyrodigal >=3.0
     - pyfastx >=0.8
     - biopython
     - numpy
     - pandas
     - psutil
-    - seqkit
     - requests
     - tqdm
 

.conda/meta.yaml

@@ -22,13 +22,12 @@ requirements:
     - poetry
     - pip
   run:
-    - hmmer >=3.3
-    - prodigal >=2.6
+    - pyhmmer >=0.10
+    - pyrodigal >=3.0
     - pyfastx >=0.8
     - biopython >=1.8
     - numpy
     - pandas
-    - seqkit
     - psutil >=5.9
     - requests
     - tqdm

.github/workflows/tests.yml

@@ -7,49 +7,47 @@ on:
       - '**.ipynb'
       - '**.bib'
       - 'ms/*'
+  # No paths-ignore here: these jobs are required status checks, and a PR whose
+  # only changes match an ignore pattern would otherwise skip them and deadlock
+  # at "Expected - waiting for status to be reported".
   pull_request:
-    types: [opened, reopened, edited]
-    paths-ignore:
-      - '**.md'
-      - '**.ipynb'
-      - '**.bib'
-      - 'ms/*'
+    types: [opened, reopened, edited, synchronize]
 
 jobs:
 
-  create-env:
-    name: ${{ matrix.os }} 
+  test:
+    name: ${{ matrix.os }} / py${{ matrix.python-version }}
     runs-on: ${{ matrix.os }}
-    defaults:
-      run:
-        shell: bash -l {0}
 
     strategy:
       fail-fast: false
       matrix:
         os: [ubuntu-latest, macos-latest]
+        python-version: ["3.9", "3.12"]
 
     steps:
       - name: checkout repository
-        uses: actions/checkout@v3
+        uses: actions/checkout@v4
         with:
           fetch-depth: 0
 
-      - name: create environment
-        uses: conda-incubator/setup-miniconda@v2
+      - name: Set up Python ${{ matrix.python-version }}
+        uses: actions/setup-python@v5
         with:
-          python-version: 3.8
-          # mamba-version: "*"
-          channels: conda-forge,bioconda,defaults
-          auto-activate-base: false
-          activate-environment: tests_pynteny
-          environment-file: envs/pynteny-dev.yml
+          python-version: ${{ matrix.python-version }}
 
-      - name: Build & Install Pynteny
-        run: poetry build && pip install dist/pynteny*.whl
+      # Pynteny is a pure-pip package (no conda / external binaries required).
+      - name: Install Pynteny
+        run: |
+          python -m pip install --upgrade pip
+          python -m pip install .
+          python -m pip install coverage
 
       - name: Run tests and collect coverage
-        run: pynteny --help && coverage run --omit pynteny/wrappers.py,pynteny/cli.py,pynteny/subcommands.py,pynteny/utils.py -m unittest discover tests && coverage xml
+        run: |
+          pynteny --help
+          coverage run --omit pynteny/wrappers.py,pynteny/cli.py,pynteny/subcommands.py,pynteny/utils.py -m unittest discover tests
+          coverage xml
 
       - name: Upload coverage to Codecov
         uses: codecov/codecov-action@v3

CONTRIBUTING.md

@@ -32,36 +32,30 @@ Unsure where to begin contributing to Pynteny? You can start by looking for issu
 
 ## Setting up a local developer environment
 
-Pynteny depends on packages that are not available in pip, namely [HMMER](https://github.com/EddyRivasLab/hmmer) and [Prodigal](https://github.com/hyattpd/Prodigal). These can be installed from the bioconda channel. Hence, to setup up a developer environment for Pynteny:
+Pynteny is a pure-Python package. Its dependencies (including the HMMER and
+Prodigal engines, provided by [pyhmmer](https://github.com/althonos/pyhmmer)
+and [pyrodigal](https://github.com/althonos/pyrodigal)) are all installable
+from pip, so no conda or external binaries are required. To set up a developer
+environment for Pynteny:
 
 1. Fork and download repo, cd to downloaded directory. You should create a new branch to work on your issue.
 
-2. Create conda environment with required dependencies:
-
-The file `envs/pynteny-dev.yml` contains all dependencies required to use Pynteny. Conda is very slow solving the environment. It is recommended to use [mamba](https://github.com/mamba-org/mamba) instead:
-
-```bash
-mamba env create -n pynteny-dev -f envs/pynteny-dev.yml
-conda activate pynteny-dev
-```
-
-3. Build package
+2. Create a virtual environment and install Pynteny in editable mode:
 
 ```bash
-(pynteny-dev) poetry build
+python -m venv .venv
+source .venv/bin/activate
+pip install -e .
 ```
 
-4. Install Pynteny
+3. Run tests
 
 ```bash
-(pynteny-dev) pip install dist/pynteny*.whl
+python -m unittest discover tests
 ```
 
-5. Run tests
-
-```bash
-(pynteny-dev) python -m unittest discover tests
-```
+A conda environment file (`envs/pynteny-dev.yml`) is still provided for
+contributors who prefer conda, but it is no longer required.
 
 ## Using GitHub codespaces
 

Dockerfile

@@ -1,27 +1,16 @@
-# FROM mcr.microsoft.com/devcontainers/miniconda:0-3
-FROM mambaorg/micromamba:latest
-USER root
+FROM python:3.11-slim
 
 WORKDIR /Pynteny
 # Copy repo to docker container
 COPY src/pynteny src/pynteny/
 COPY tests tests/
-COPY envs envs/
 COPY README.md .
 COPY pyproject.toml .
 COPY LICENSE .
 
-# Make conda environment and activate
-# RUN conda install mamba -n base -c conda-forge
-RUN micromamba env create -f envs/pynteny-dev.yml
-SHELL ["micromamba", "run", "-n", "pynteny-dev", "/bin/bash", "-c"]
-# Build and install Pynteny
-RUN poetry build && pip install dist/pynteny*.whl && pynteny --help
-# Give read/write permissions to install directory (needed to make config.json)
-RUN chmod ugo+rw /opt/conda/envs/pynteny-dev/lib/python3.10/site-packages
-
-# Initialize conda for default user
-RUN micromamba shell init --shell=bash
-# # Activate pynteny environment by default
-RUN echo "micromamba activate pynteny-dev" >> ../root/.bashrc
-RUN source ../root/.bashrc
+# Build and install Pynteny (pure pip, no conda required).
+# pyhmmer and pyrodigal ship manylinux wheels, so no compiler toolchain is needed.
+RUN pip install --no-cache-dir poetry \
+    && poetry build \
+    && pip install --no-cache-dir dist/pynteny*.whl \
+    && pynteny --help

README.md

@@ -36,35 +36,37 @@ Check out the [Pynteny paper](https://doi.org/10.21105/joss.05289) in the Journa
 
 ## 2. :wrench: Setup
 
-Install with conda:
+Pynteny is a pure-Python package (it no longer requires conda or any external
+binaries). All dependencies, including the HMMER and Prodigal engines, are
+provided by the pip packages [pyhmmer](https://github.com/althonos/pyhmmer) and
+[pyrodigal](https://github.com/althonos/pyrodigal).
 
-1. Pynteny requires Python 3.10. The easiest way to handle dependencies is by creating a dedicated conda environment:
+Install with pip:
 
 ```bash
-conda create -n pynteny -c bioconda -c conda-forge python=3.10 pynteny
-conda activate pynteny
+pip install pynteny
 ```
 
-2. Check that installation worked fine:
+Or with conda (from bioconda):
 
 ```bash
-(pynteny) pynteny --help
+conda create -n pynteny -c bioconda -c conda-forge python=3.10 pynteny
+conda activate pynteny
 ```
-### 2.1. Installing on Windows
 
-Pynteny is designed to run on Linux machines. However, it can be installed within the [Windows Subsystem for Linux](https://learn.microsoft.com/en-us/windows/wsl/install) via conda.
-
-### 2.2. Installing on MacOS with the latest ARM64 architecture
-
-Pynteny doesn't currently support the latest ARM64 architecture of silicon processors (e.g. MacBook M1 and M2). If that is your case, you can install Pynteny using the workaround below (based on [this post](https://towardsdatascience.com/how-to-manage-conda-environments-on-an-apple-silicon-m1-mac-1e29cb3bad12)):
+Check that the installation worked fine:
 
 ```bash
-CONDA_SUBDIR=osx-64 conda create -n pynteny_x86 python=3.10
-conda activate pynteny_x86
-conda config --env --set subdir osx-64
-conda install -c bioconda pynteny
+pynteny --help
 ```
 
+### 2.1. Installing on Windows
+
+Pynteny is developed and tested on Linux, but since it is now a pure-Python
+package, `pip install pynteny` also works on Windows and macOS (including Apple
+Silicon / ARM64), provided wheels are available for pyhmmer and pyrodigal on
+your platform.
+
 ## 3. :rocket: Usage
 
 Consider the following toy example of a syntenic block:
@@ -126,11 +128,10 @@ You can find more notebooks in the [examples directory](docs/examples/). Find mo
 ## 5. :arrows_counterclockwise: Dependencies
 Pynteny would not work without these awesome projects:
 
-- [hmmer](https://github.com/EddyRivasLab/hmmer)
-- [prodigal](https://github.com/hyattpd/Prodigal)
+- [pyhmmer](https://github.com/althonos/pyhmmer)
+- [pyrodigal](https://github.com/althonos/pyrodigal)
 - [pyfastx](https://github.com/lmdu/pyfastx)
 - [biopython](https://github.com/biopython/biopython)
-- [seqkit](https://bioinf.shenwei.me/seqkit/)
 - [numpy](https://github.com/numpy/numpy)
 - [pandas](https://github.com/pandas-dev/pandas)
 - [psutil](https://github.com/giampaolo/psutil)

docs/examples/example_api_colab.ipynb

@@ -41,33 +41,14 @@
       "outputs": [],
       "source": [
         "%%capture\n",
-        "#@title Install Pynteny (can take a few minutes)\n",
+        "#@title Install Pynteny\n",
         "try:\n",
         "  import google.colab\n",
-        "  import sys\n",
-        "  # Install miniconda\n",
-        "  !wget -qnc https://repo.anaconda.com/miniconda/Miniconda3-py38_22.11.1-1-Linux-x86_64.sh\n",
-        "  !chmod +x Miniconda3-py38_22.11.1-1-Linux-x86_64.sh\n",
-        "  !bash ./Miniconda3-py38_22.11.1-1-Linux-x86_64.sh -bfp /usr/local/ 2>&1 1>/dev/null\n",
-        "  !rm ./Miniconda3-py38_22.11.1-1-Linux-x86_64.sh\n",
-        "  # Install Pynteny and dependencies\n",
-        "  !conda install -q -y -c conda-forge -c -c bioconda \\\n",
-        "      python=3.8 \\\n",
-        "      poetry=1.3 \\\n",
-        "      pip \\\n",
-        "      hmmer=3.3 \\\n",
-        "      prodigal=2.6 \\\n",
-        "      pyfastx=0.8 \\\n",
-        "      biopython=1.80 \\\n",
-        "      psutil=5.9 \\\n",
-        "      tqdm \\\n",
-        "      requests 2>&1 1>/dev/null\n",
-        "  !pip install --upgrade numpy==1.21.6\n",
-        "  !pip install --upgrade pandas==1.3.5\n",
-        "  !git clone https://github.com/Robaina/Pynteny.git\n",
-        "  !cd Pynteny && poetry build && pip install -q dist/pynteny*.whl\n",
-        "  !cp -r Pynteny/tests/ . && rm -r Pynteny\n",
-        "  sys.path.insert(0, f\"/usr/local/lib/python3.8/site-packages/\")\n",
+        "  # Pynteny is a pure-pip package (no conda / external binaries required).\n",
+        "  !pip install -q pynteny\n",
+        "  # Grab the bundled test data used throughout this example.\n",
+        "  !git clone -q https://github.com/Robaina/Pynteny.git\n",
+        "  !cp -r Pynteny/tests/ . && rm -rf Pynteny\n",
         "except ImportError:\n",
         "  pass"
       ]

docs/index.md

@@ -13,9 +13,18 @@ These are the available subcommands in the CLI, run as ```pynteny <subcommand> <
 
 ## 2. Setup
 
-Install with conda:
+Pynteny is a pure-Python package (it no longer requires conda or any external
+binaries). The HMMER and Prodigal engines are provided by the pip packages
+[pyhmmer](https://github.com/althonos/pyhmmer) and
+[pyrodigal](https://github.com/althonos/pyrodigal).
 
-Pynteny requires Python 3.10. The easiest way to handle dependencies is by creating a dedicated conda environment:
+Install with pip:
+
+```bash
+pip install pynteny
+```
+
+Or with conda (from bioconda):
 
 ```bash
 conda create -n pynteny -c bioconda -c conda-forge python=3.10 pynteny
@@ -25,22 +34,15 @@ conda activate pynteny
 Check that installation worked fine:
 
 ```bash
-(pynteny) pynteny --help
+pynteny --help
 ```
-### 2.1. Installing on Windows
 
-Pynteny is designed to run on Linux machines. However, it can be installed within the [Windows Subsystem for Linux](https://learn.microsoft.com/en-us/windows/wsl/install) via conda.
+### 2.1. Installing on Windows and macOS
 
-### 2.2. Installing on MacOS with the latest ARM64 architecture
-
-Pynteny doesn't currently support the latest ARM64 architecture of silicon processors (e.g. MacBook M1 and M2). If that is your case, you can install Pynteny using the workaround below (based on [this post](https://towardsdatascience.com/how-to-manage-conda-environments-on-an-apple-silicon-m1-mac-1e29cb3bad12)):
-
-```bash
-CONDA_SUBDIR=osx-64 conda create -n pynteny_x86 python=3.10
-conda activate pynteny_x86
-conda config --env --set subdir osx-64
-conda install -c bioconda pynteny
-```
+Pynteny is developed and tested on Linux, but since it is now a pure-Python
+package, `pip install pynteny` also works on Windows and macOS (including Apple
+Silicon / ARM64), provided wheels are available for pyhmmer and pyrodigal on
+your platform.
 
 ## 3. General usage
 

envs/pynteny-dev.yml

@@ -7,17 +7,16 @@ dependencies:
   - python >= 3.8
   - poetry >= 1.3
   - pip
-  - hmmer >= 3.3
-  - prodigal >= 2.6
   - pyfastx >=0.8
   - biopython >= 1.8
   - numpy
   - pandas
-  - seqkit
   - psutil >= 5.9
   - requests
   - tqdm
   - pip:
+    - pyhmmer >=0.10
+    - pyrodigal >=3.0
     - mkdocs
     - mkdocs-gen-files
     - pymdown-extensions

pyproject.toml

@@ -24,6 +24,15 @@ packages = [{ include = "pynteny", from = "src" }]
 include = ["src/pynteny/config.json"]
 [tool.poetry.dependencies]
 python = "^3.8"
+pyhmmer = ">=0.10"
+pyrodigal = ">=3.0"
+pyfastx = ">=0.8"
+biopython = ">=1.8"
+numpy = "*"
+pandas = "*"
+psutil = ">=5.9"
+requests = "*"
+tqdm = "*"
 [tool.poetry.scripts]
 pynteny = "pynteny.cli:main"
 

src/pynteny/filter.py

@@ -220,9 +220,9 @@ def _assign_code_to_HMM(self, hmm_group_name: str) -> int:
     def _add_meta_codes_to_HMM_hits(self, all_hit_labels: pd.Dataframe) -> pd.Dataframe:
         """Add numeric codes for each hmm and strand, compute distance between genes"""
         all_hit_labels["gene_pos_diff"] = all_hit_labels.gene_pos.diff()
-        all_hit_labels.loc[
-            0, "gene_pos_diff"
-        ] = 1  # required for rolling (skips first nan)
+        all_hit_labels.loc[0, "gene_pos_diff"] = (
+            1  # required for rolling (skips first nan)
+        )
         all_hit_labels["hmm_code"] = all_hit_labels.hmm.apply(self._assign_code_to_HMM)
         all_hit_labels["strand"] = all_hit_labels.strand.apply(
             lambda strand: -1 if strand == "neg" else (1 if strand == "pos" else 0)