# Base Image
FROM python:3.8.20-bullseye

# Install required system dependencies and clean up apt cache
RUN apt-get update && apt-get install -y \
    wget \
    cmake \
    g++ \
    make \
    zlib1g-dev \
    libjemalloc-dev \
    unzip \
    git \
    ncbi-blast+ \
    && rm -rf /var/lib/apt/lists/*  # Clean apt cache to reduce image size

# Install Python dependencies
RUN pip install biopython==1.75 ete3==3.1.2 tqdm==4.66.4 six==1.17.0

# Clone EukPhylo repository
RUN git clone -b Docker https://github.com/Katzlab/EukPhylo.git

# Install VSEARCH
RUN wget https://github.com/torognes/vsearch/releases/download/v2.21.0/vsearch-2.21.0-linux-aarch64.tar.gz && \
    tar xzf vsearch-2.21.0-linux-aarch64.tar.gz && \
    mv vsearch-2.21.0-linux-aarch64 /usr/local/vsearch && \
    ln -s /usr/local/vsearch/bin/vsearch /usr/local/bin/vsearch && \
    rm -rf vsearch-2.21.0-linux-aarch64.tar.gz

# Compile DIAMOND v2.1.11 from source
RUN wget https://github.com/bbuchfink/diamond/archive/refs/tags/v2.1.11.tar.gz && \
    tar xzf v2.1.11.tar.gz && \
    cd diamond-2.1.11 && \
    mkdir build && cd build && \
    cmake .. && \
    make -j$(nproc) && \
    cp diamond /usr/local/bin/ && \
    cd ../.. && rm -rf diamond-2.1.11 v2.1.11.tar.gz

# Verify DIAMOND installation
RUN diamond --version

# Verify installations
RUN diamond --version && vsearch --version && blastn -version

# Ensure executables are accessible
ENV PATH="$PATH:/usr/local/vsearch/bin:/usr/local/bin"

# Change executable permissions on wrapper script
RUN chmod +x /EukPhylo/PTL1/Transcriptomes/Scripts/wrapper_submit.sh

# Set container entrypoint
ENTRYPOINT ["bash", "/EukPhylo/PTL1/Transcriptomes/Scripts/wrapper_submit.sh"]