From 2a9377bdfeb93cd770cbd22bc6f9715a7d150da0 Mon Sep 17 00:00:00 2001
From: baichom1 <baichom1@mskcc.org>
Date: Fri, 10 Apr 2026 10:57:31 -0400
Subject: [PATCH 01/17] try claudes attempt at improving this dockerfile

---
 biobakery-profiler/Dockerfile                 | 133 +++++++++++-------
 .../config_4.0.0.alpha.1-final-smaller.patch  |  14 ++
 2 files changed, 98 insertions(+), 49 deletions(-)
 create mode 100644 biobakery-profiler/config_4.0.0.alpha.1-final-smaller.patch

diff --git a/biobakery-profiler/Dockerfile b/biobakery-profiler/Dockerfile
index 6356888..3aea350 100644
--- a/biobakery-profiler/Dockerfile
+++ b/biobakery-profiler/Dockerfile
@@ -1,65 +1,100 @@
-FROM python:3.9.17-bookworm AS dbbuild
+# ============================================================
+# Stage 1: Build samtools
+# ============================================================
+FROM python:3.9.17-slim-bookworm AS dbbuild
+
 RUN apt-get update && \
-    DEBIAN_FRONTEND="noninteractive" apt-get install --no-install-recommends -y wget  ca-certificates build-essential  zlib1g-dev  libbz2-dev liblzma-dev && \
+    DEBIAN_FRONTEND="noninteractive" apt-get install --no-install-recommends -y \
+        wget ca-certificates build-essential zlib1g-dev libbz2-dev liblzma-dev && \
     rm -rf /var/lib/apt/lists/*
 
-# RUN mkdir -p /dbs/util/ && \
-#    wget  -c --read-timeout=5 http://huttenhower.sph.harvard.edu/humann_data/full_mapping_v201901b.tar.gz && \
-#    cd /dbs/util/ && \
-#    tar xzf /full_mapping_v201901b.tar.gz && \
-#    rm /full_mapping_v201901b.tar.gz
-RUN wget https://github.com/samtools/samtools/releases/download/1.15.1/samtools-1.15.1.tar.bz2 && \
-    tar xjf samtools-1.15.1.tar.bz2  && cd samtools-1.15.1 && \
-    ./configure --without-curses && make && make install && which samtools #
-
-
-
-#FROM ubuntu:18.04 AS runtime-image
-FROM python:3.9.17-bookworm  AS runtime-image
+RUN wget -q https://github.com/samtools/samtools/releases/download/1.15.1/samtools-1.15.1.tar.bz2 && \
+    tar xjf samtools-1.15.1.tar.bz2 && \
+    cd samtools-1.15.1 && \
+    ./configure --without-curses && \
+    make -j$(nproc) && \
+    make install && \
+    which samtools
 
+# ============================================================
+# Stage 2: Build Python packages in a venv we can copy
+# ============================================================
+FROM python:3.9.17-slim-bookworm AS pybuild
 
 ARG HUMANNVERSION=0.0.0
 ARG METAPHLANVERSION=0.0.0
 
-COPY --from=dbbuild /usr/local/bin/samtools /usr/local/bin/samtools
-
-WORKDIR /tmp
-RUN export LC_ALL=en_US.UTF-8 && export LANG=en_US.UTF-8
-
 RUN apt-get update && \
-    DEBIAN_FRONTEND="noninteractive" apt-get install --no-install-recommends -y  build-essential   libjpeg-dev zlib1g-dev libbz2-dev liblzma-dev ncbi-blast+ mafft raxml && \
+    DEBIAN_FRONTEND="noninteractive" apt-get install --no-install-recommends -y \
+        build-essential \
+        wget \
+        ca-certificates \
+        patch \
+        libjpeg-dev \
+        zlib1g-dev \
+        libbz2-dev \
+        liblzma-dev && \
     rm -rf /var/lib/apt/lists/*
 
+RUN python3 -m venv /opt/venv
+ENV PATH="/opt/venv/bin:$PATH"
+
+COPY nucleotide_${HUMANNVERSION}.patch /tmp/nucleotide.patch
+COPY config_${HUMANNVERSION}.patch /tmp/config.patch
+
+RUN pip install --no-cache-dir numpy cython && \
+    pip install --no-cache-dir boto3 cloudpickle awscli && \
+    pip install --no-cache-dir biom-format && \
+    wget -qO- https://github.com/biobakery/humann/archive/refs/tags/${HUMANNVERSION}.tar.gz \
+        | tar xz -C /tmp && \
+    mv /tmp/humann-* /tmp/humann && \
+    cd /tmp/humann && \
+    patch humann/search/nucleotide.py < /tmp/nucleotide.patch && \
+    patch humann/config.py < /tmp/config.patch && \
+    mv setup.py tmp.py && \
+    sed 's|2\.2\.3|2\.5\.1|g' tmp.py > tmp2.py && \
+    sed 's|bowtie2_folder="bowtie2-2\.5\.1|bowtie2_folder="bowtie2-2\.5\.1-linux-x86_64|g' tmp2.py > setup.py && \
+    pip install --no-cache-dir . --no-binary :all: && \
+    pip install --no-cache-dir metaphlan==${METAPHLANVERSION} && \
+    find /opt/venv -type d -name "__pycache__" -exec rm -rf {} + 2>/dev/null; \
+    find /opt/venv -type d -name "tests" -exec rm -rf {} + 2>/dev/null; \
+    find /opt/venv -type d -name "test" -exec rm -rf {} + 2>/dev/null; \
+    find /opt/venv -name "*.pyc" -delete 2>/dev/null; \
+    find /opt/venv -name "*.pyo" -delete 2>/dev/null; \
+    rm -rf /tmp/humann /tmp/*.patch
+
+# ============================================================
+# Stage 3: Minimal runtime image
+# ============================================================
+FROM python:3.9.17-slim-bookworm AS runtime-image
 
-# You may be tempted to do these all in one call but unfortunately
-# (eg) biom-format wont build without numpy, and it wont install it automatically
-
-# see the glpsol issue here https://forum.biobakery.org/t/silent-errors-with-bad-install-of-libglpk/4814
-
-# this nucleotide.py makes bowtie indexing parallel
-ADD nucleotide_${HUMANNVERSION}.patch /tmp/nucleotide.patch
-ADD config_${HUMANNVERSION}.patch /tmp/config.patch
+ARG HUMANNVERSION=0.0.0
+ARG METAPHLANVERSION=0.0.0
 
-# We also have to bump to a more recent version of bowtie that is multithreaded,
-# so we clumsily make that with some uninspired sed calls
-RUN pip3 install boto3 cloudpickle awscli && \
-    git clone --depth=1 --branch ${HUMANNVERSION} https://github.com/biobakery/humann.git humann && \
-    du -h --max-depth=0 humann && \
-    cd humann && \
-    patch  humann/search/nucleotide.py  < /tmp/nucleotide.patch && \
-    patch  humann/config.py  < /tmp/config.patch && \
-    mv setup.py tmp.py && cat tmp.py | sed "s|2\.2\.3|2\.5\.1|g" > tmp2.py  && \
-    sed "s|bowtie2_folder\=\"bowtie2-2\.5\.1|bowtie2_folder\=\"bowtie2-2\.5\.1-linux-x86_64|g" tmp2.py > setup.py && \
-    pip3 install -vvv . --no-binary :all: && \
-    ldconfig /usr/local/lib && glpsol --version && \
-    pip3 install numpy cython && pip3 install --no-cache-dir biom-format && \
-    pip3 install metaphlan==${METAPHLANVERSION} &&  du --max-depth 1 -h  ~/.cache/pip && \
-    rm -r  ~/.cache/pip
+# Copy samtools binary and its needed libs
+COPY --from=dbbuild /usr/local/bin/samtools /usr/local/bin/samtools
 
-# The --help is so you can check to ensure that the grouping table are properly read
-#RUN humann_config --update database_folders utility_mapping /dbs/util/ && \
-#    humann_regroup_table --help
+# Copy the entire pre-built venv
+COPY --from=pybuild /opt/venv /opt/venv
+ENV PATH="/opt/venv/bin:$PATH"
 
-#RUN humann_test --run-functional-tests-end-to-end
+# Install ONLY runtime dependencies — no build-essential, no git
+RUN apt-get update && \
+    DEBIAN_FRONTEND="noninteractive" apt-get install --no-install-recommends -y \
+        libjpeg62-turbo \
+        zlib1g \
+        libbz2-1.0 \
+        liblzma5 \
+        ncbi-blast+ \
+        mafft \
+        raxml \
+        libgomp1 \
+        libglpk40 && \
+    rm -rf /var/lib/apt/lists/* && \
+    ldconfig /usr/local/lib && \
+    glpsol --version
+
+ENV LC_ALL=en_US.UTF-8
+ENV LANG=en_US.UTF-8
 
-#RUN cd humann-3.9 && ls examples &&  humann --input examples/demo.fasta.gz --output tmpout
+WORKDIR /tmp
diff --git a/biobakery-profiler/config_4.0.0.alpha.1-final-smaller.patch b/biobakery-profiler/config_4.0.0.alpha.1-final-smaller.patch
new file mode 100644
index 0000000..c5246b8
--- /dev/null
+++ b/biobakery-profiler/config_4.0.0.alpha.1-final-smaller.patch
@@ -0,0 +1,14 @@
+105c105,107
+< user_edit_config_file="humann.cfg"
+---
+> # User config file
+> if not os.environ.get("HUMANN_CONFIG"):
+>     user_edit_config_file="humann.cfg"
+107,108c109,112
+< full_path_user_edit_config_file=os.path.join(os.path.dirname(os.path.abspath(__file__)),
+<     user_edit_config_file)
+---
+>     full_path_user_edit_config_file=os.path.join(os.path.dirname(os.path.abspath(__file__)),
+>                                                  user_edit_config_file)
+> else:
+>     full_path_user_edit_config_file=os.environ.get("HUMANN_CONFIG")

From 74cce6e20dc7b6b8bbff99c6cc2bb7e1770f62a9 Mon Sep 17 00:00:00 2001
From: baichom1 <baichom1@mskcc.org>
Date: Fri, 10 Apr 2026 11:02:08 -0400
Subject: [PATCH 02/17] update build manifest

---
 biobakery_build_manifest.csv | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/biobakery_build_manifest.csv b/biobakery_build_manifest.csv
index 527a0f5..bab1bb7 100644
--- a/biobakery_build_manifest.csv
+++ b/biobakery_build_manifest.csv
@@ -1,3 +1,3 @@
-biobakery-profiler,4.0.6--4.0.0.alpha.1-final
+biobakery-profiler,4.0.6--4.0.0.alpha.1-final-smaller
 biobakery-profiler,4.1.0--v3.9
 biobakery-profiler,4.0.5--3.6.1

From 85b2d311b61528bc2386e8cb6a8222d41a19c02b Mon Sep 17 00:00:00 2001
From: baichom1 <baichom1@mskcc.org>
Date: Fri, 10 Apr 2026 11:03:41 -0400
Subject: [PATCH 03/17] rename nucleotide as well

---
 ...1-final.patch => nucleotide_4.0.0.alpha.1-final-smaller.patch} | 0
 1 file changed, 0 insertions(+), 0 deletions(-)
 rename biobakery-profiler/{nucleotide_4.0.0.alpha.1-final.patch => nucleotide_4.0.0.alpha.1-final-smaller.patch} (100%)

diff --git a/biobakery-profiler/nucleotide_4.0.0.alpha.1-final.patch b/biobakery-profiler/nucleotide_4.0.0.alpha.1-final-smaller.patch
similarity index 100%
rename from biobakery-profiler/nucleotide_4.0.0.alpha.1-final.patch
rename to biobakery-profiler/nucleotide_4.0.0.alpha.1-final-smaller.patch

From 345b015d961c2c4d195318eac47a91c7087aa8a6 Mon Sep 17 00:00:00 2001
From: miraep8 <miraep8@gmail.com>
Date: Fri, 10 Apr 2026 11:08:13 -0400
Subject: [PATCH 04/17] add glpsol

---
 biobakery-profiler/Dockerfile | 3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)

diff --git a/biobakery-profiler/Dockerfile b/biobakery-profiler/Dockerfile
index 3aea350..2d6c946 100644
--- a/biobakery-profiler/Dockerfile
+++ b/biobakery-profiler/Dockerfile
@@ -89,7 +89,8 @@ RUN apt-get update && \
         mafft \
         raxml \
         libgomp1 \
-        libglpk40 && \
+        libglpk40 \
+        glpk-utils && \
     rm -rf /var/lib/apt/lists/* && \
     ldconfig /usr/local/lib && \
     glpsol --version

From 6fcd4c265b2540af0d165ad65ea1952a22997267 Mon Sep 17 00:00:00 2001
From: miraep8 <miraep8@gmail.com>
Date: Fri, 10 Apr 2026 12:23:14 -0400
Subject: [PATCH 05/17] -verify-build

---
 biobakery-profiler/Dockerfile | 39 +++++++++++++++++++++++++++++++----
 1 file changed, 35 insertions(+), 4 deletions(-)

diff --git a/biobakery-profiler/Dockerfile b/biobakery-profiler/Dockerfile
index 2d6c946..2f02fae 100644
--- a/biobakery-profiler/Dockerfile
+++ b/biobakery-profiler/Dockerfile
@@ -8,6 +8,13 @@ RUN apt-get update && \
         wget ca-certificates build-essential zlib1g-dev libbz2-dev liblzma-dev && \
     rm -rf /var/lib/apt/lists/*
 
+#old command that was at this stage, commented out:
+# RUN mkdir -p /dbs/util/ && \
+#    wget  -c --read-timeout=5 http://huttenhower.sph.harvard.edu/humann_data/full_mapping_v201901b.tar.gz && \
+#    cd /dbs/util/ && \
+#    tar xzf /full_mapping_v201901b.tar.gz && \
+#    rm /full_mapping_v201901b.tar.gz    
+
 RUN wget -q https://github.com/samtools/samtools/releases/download/1.15.1/samtools-1.15.1.tar.bz2 && \
     tar xjf samtools-1.15.1.tar.bz2 && \
     cd samtools-1.15.1 && \
@@ -42,7 +49,10 @@ ENV PATH="/opt/venv/bin:$PATH"
 COPY nucleotide_${HUMANNVERSION}.patch /tmp/nucleotide.patch
 COPY config_${HUMANNVERSION}.patch /tmp/config.patch
 
-RUN pip install --no-cache-dir numpy cython && \
+# We also have to bump to a more recent version of bowtie that is multithreaded,
+# so we clumsily make that with some uninspired sed calls
+RUN set -e && \
+    pip install --no-cache-dir numpy cython && \
     pip install --no-cache-dir boto3 cloudpickle awscli && \
     pip install --no-cache-dir biom-format && \
     wget -qO- https://github.com/biobakery/humann/archive/refs/tags/${HUMANNVERSION}.tar.gz \
@@ -51,10 +61,19 @@ RUN pip install --no-cache-dir numpy cython && \
     cd /tmp/humann && \
     patch humann/search/nucleotide.py < /tmp/nucleotide.patch && \
     patch humann/config.py < /tmp/config.patch && \
-    mv setup.py tmp.py && \
-    sed 's|2\.2\.3|2\.5\.1|g' tmp.py > tmp2.py && \
-    sed 's|bowtie2_folder="bowtie2-2\.5\.1|bowtie2_folder="bowtie2-2\.5\.1-linux-x86_64|g' tmp2.py > setup.py && \
+    # --- Safer setup.py modification with verification ---
+    cp setup.py setup.py.orig && \
+    sed -i 's|2\.2\.3|2\.5\.1|g' setup.py && \
+    sed -i 's|bowtie2_folder="bowtie2-2\.5\.1|bowtie2_folder="bowtie2-2\.5\.1-linux-x86_64|g' setup.py && \
+    echo "=== setup.py diff ===" && \
+    diff setup.py.orig setup.py || true && \
+    echo "=== Installing humann ===" && \
     pip install --no-cache-dir . --no-binary :all: && \
+    # --- Verify it actually installed ---
+    echo "=== Verifying humann install ===" && \
+    which humann_config && \
+    humann --version && \
+    ls /opt/venv/bin/humann* && \
     pip install --no-cache-dir metaphlan==${METAPHLANVERSION} && \
     find /opt/venv -type d -name "__pycache__" -exec rm -rf {} + 2>/dev/null; \
     find /opt/venv -type d -name "tests" -exec rm -rf {} + 2>/dev/null; \
@@ -78,6 +97,10 @@ COPY --from=dbbuild /usr/local/bin/samtools /usr/local/bin/samtools
 COPY --from=pybuild /opt/venv /opt/venv
 ENV PATH="/opt/venv/bin:$PATH"
 
+# You may be tempted to do these all in one call but unfortunately
+# (eg) biom-format wont build without numpy, and it wont install it automatically
+
+# see the glpsol issue here https://forum.biobakery.org/t/silent-errors-with-bad-install-of-libglpk/4814
 # Install ONLY runtime dependencies — no build-essential, no git
 RUN apt-get update && \
     DEBIAN_FRONTEND="noninteractive" apt-get install --no-install-recommends -y \
@@ -99,3 +122,11 @@ ENV LC_ALL=en_US.UTF-8
 ENV LANG=en_US.UTF-8
 
 WORKDIR /tmp
+
+# The --help is so you can check to ensure that the grouping table are properly read
+#RUN humann_config --update database_folders utility_mapping /dbs/util/ && \
+#    humann_regroup_table --help
+
+#RUN humann_test --run-functional-tests-end-to-end
+
+#RUN cd humann-3.9 && ls examples &&  humann --input examples/demo.fasta.gz --output tmpout

From 7fdcdfa96c69796303ffaed0c32b3ce3ec3b180f Mon Sep 17 00:00:00 2001
From: miraep8 <miraep8@gmail.com>
Date: Fri, 10 Apr 2026 12:26:43 -0400
Subject: [PATCH 06/17] bump v

---
 ...maller.patch => config_4.0.0.alpha.1-final-smaller-v2.patch} | 0
 ...er.patch => nucleotide_4.0.0.alpha.1-final-smaller-v2.patch} | 0
 biobakery_build_manifest.csv                                    | 2 +-
 3 files changed, 1 insertion(+), 1 deletion(-)
 rename biobakery-profiler/{config_4.0.0.alpha.1-final-smaller.patch => config_4.0.0.alpha.1-final-smaller-v2.patch} (100%)
 rename biobakery-profiler/{nucleotide_4.0.0.alpha.1-final-smaller.patch => nucleotide_4.0.0.alpha.1-final-smaller-v2.patch} (100%)

diff --git a/biobakery-profiler/config_4.0.0.alpha.1-final-smaller.patch b/biobakery-profiler/config_4.0.0.alpha.1-final-smaller-v2.patch
similarity index 100%
rename from biobakery-profiler/config_4.0.0.alpha.1-final-smaller.patch
rename to biobakery-profiler/config_4.0.0.alpha.1-final-smaller-v2.patch
diff --git a/biobakery-profiler/nucleotide_4.0.0.alpha.1-final-smaller.patch b/biobakery-profiler/nucleotide_4.0.0.alpha.1-final-smaller-v2.patch
similarity index 100%
rename from biobakery-profiler/nucleotide_4.0.0.alpha.1-final-smaller.patch
rename to biobakery-profiler/nucleotide_4.0.0.alpha.1-final-smaller-v2.patch
diff --git a/biobakery_build_manifest.csv b/biobakery_build_manifest.csv
index bab1bb7..9cc97e4 100644
--- a/biobakery_build_manifest.csv
+++ b/biobakery_build_manifest.csv
@@ -1,3 +1,3 @@
-biobakery-profiler,4.0.6--4.0.0.alpha.1-final-smaller
+biobakery-profiler,4.0.6--4.0.0.alpha.1-final-smaller-v2
 biobakery-profiler,4.1.0--v3.9
 biobakery-profiler,4.0.5--3.6.1

From ce0be69321b5f3e3c68cef97d942cdac0a0aafbc Mon Sep 17 00:00:00 2001
From: miraep8 <miraep8@gmail.com>
Date: Fri, 10 Apr 2026 12:59:42 -0400
Subject: [PATCH 07/17] more intense debugging

---
 biobakery-profiler/Dockerfile                 | 76 ++++++++++++-------
 ...alpha.1-final-smaller-v2-with_tests.patch} |  0
 ...alpha.1-final-smaller-v2-with_tests.patch} |  0
 biobakery_build_manifest.csv                  |  2 +-
 4 files changed, 51 insertions(+), 27 deletions(-)
 rename biobakery-profiler/{config_4.0.0.alpha.1-final-smaller-v2.patch => config_4.0.0.alpha.1-final-smaller-v2-with_tests.patch} (100%)
 rename biobakery-profiler/{nucleotide_4.0.0.alpha.1-final-smaller-v2.patch => nucleotide_4.0.0.alpha.1-final-smaller-v2-with_tests.patch} (100%)

diff --git a/biobakery-profiler/Dockerfile b/biobakery-profiler/Dockerfile
index 2f02fae..592de00 100644
--- a/biobakery-profiler/Dockerfile
+++ b/biobakery-profiler/Dockerfile
@@ -26,6 +26,7 @@ RUN wget -q https://github.com/samtools/samtools/releases/download/1.15.1/samtoo
 # ============================================================
 # Stage 2: Build Python packages in a venv we can copy
 # ============================================================
+
 FROM python:3.9.17-slim-bookworm AS pybuild
 
 ARG HUMANNVERSION=0.0.0
@@ -49,39 +50,62 @@ ENV PATH="/opt/venv/bin:$PATH"
 COPY nucleotide_${HUMANNVERSION}.patch /tmp/nucleotide.patch
 COPY config_${HUMANNVERSION}.patch /tmp/config.patch
 
-# We also have to bump to a more recent version of bowtie that is multithreaded,
-# so we clumsily make that with some uninspired sed calls
-RUN set -e && \
-    pip install --no-cache-dir numpy cython && \
+# Install prerequisites
+RUN pip install --no-cache-dir numpy cython && \
     pip install --no-cache-dir boto3 cloudpickle awscli && \
-    pip install --no-cache-dir biom-format && \
-    wget -qO- https://github.com/biobakery/humann/archive/refs/tags/${HUMANNVERSION}.tar.gz \
-        | tar xz -C /tmp && \
+    pip install --no-cache-dir biom-format
+
+# Download and inspect humann — STOP HERE to debug
+RUN set -e && \
+    echo "=== HUMANNVERSION is: ${HUMANNVERSION} ===" && \
+    wget -qO /tmp/humann.tar.gz https://github.com/biobakery/humann/archive/refs/tags/${HUMANNVERSION}.tar.gz && \
+    echo "=== Tarball contents (top-level) ===" && \
+    tar tzf /tmp/humann.tar.gz | head -20 && \
+    tar xzf /tmp/humann.tar.gz -C /tmp && \
+    echo "=== Directories in /tmp ===" && \
+    ls -la /tmp/ && \
+    echo "=== Moving to /tmp/humann ===" && \
     mv /tmp/humann-* /tmp/humann && \
+    ls -la /tmp/humann/ && \
+    echo "=== setup.py exists? ===" && \
+    cat /tmp/humann/setup.py | head -30
+
+# Patch and modify setup.py
+RUN set -e && \
     cd /tmp/humann && \
+    echo "=== Applying patches ===" && \
     patch humann/search/nucleotide.py < /tmp/nucleotide.patch && \
     patch humann/config.py < /tmp/config.patch && \
-    # --- Safer setup.py modification with verification ---
     cp setup.py setup.py.orig && \
     sed -i 's|2\.2\.3|2\.5\.1|g' setup.py && \
     sed -i 's|bowtie2_folder="bowtie2-2\.5\.1|bowtie2_folder="bowtie2-2\.5\.1-linux-x86_64|g' setup.py && \
     echo "=== setup.py diff ===" && \
     diff setup.py.orig setup.py || true && \
-    echo "=== Installing humann ===" && \
-    pip install --no-cache-dir . --no-binary :all: && \
-    # --- Verify it actually installed ---
-    echo "=== Verifying humann install ===" && \
-    which humann_config && \
-    humann --version && \
-    ls /opt/venv/bin/humann* && \
-    pip install --no-cache-dir metaphlan==${METAPHLANVERSION} && \
-    find /opt/venv -type d -name "__pycache__" -exec rm -rf {} + 2>/dev/null; \
-    find /opt/venv -type d -name "tests" -exec rm -rf {} + 2>/dev/null; \
-    find /opt/venv -type d -name "test" -exec rm -rf {} + 2>/dev/null; \
-    find /opt/venv -name "*.pyc" -delete 2>/dev/null; \
-    find /opt/venv -name "*.pyo" -delete 2>/dev/null; \
-    rm -rf /tmp/humann /tmp/*.patch
+    echo "=== Full setup.py ===" && \
+    cat setup.py
 
+# Install humann — this is the step we need to watch carefully
+RUN set -e && \
+    echo "=== which pip ===" && \
+    which pip && \
+    echo "=== which python ===" && \
+    which python && \
+    echo "=== pip location ===" && \
+    pip --version && \
+    cd /tmp/humann && \
+    echo "=== Installing humann ===" && \
+    pip install --no-cache-dir . --no-binary :all: 2>&1 && \
+    echo "=== pip install exit code: $? ===" && \
+    echo "=== Checking install ===" && \
+    pip show humann && \
+    echo "=== venv bin contents ===" && \
+    ls -la /opt/venv/bin/ && \
+    echo "=== humann files in venv ===" && \
+    ls /opt/venv/bin/humann* || echo "NO HUMANN BINARIES FOUND" && \
+    echo "=== entry_points check ===" && \
+    pip show -f humann | head -50 && \
+    echo "=== system-wide check ===" && \
+    find / -name "humann*" -type f 2>/dev/null | head -20
 # ============================================================
 # Stage 3: Minimal runtime image
 # ============================================================
@@ -124,9 +148,9 @@ ENV LANG=en_US.UTF-8
 WORKDIR /tmp
 
 # The --help is so you can check to ensure that the grouping table are properly read
-#RUN humann_config --update database_folders utility_mapping /dbs/util/ && \
-#    humann_regroup_table --help
+RUN humann_config --update database_folders utility_mapping /dbs/util/ && \
+    humann_regroup_table --help
 
-#RUN humann_test --run-functional-tests-end-to-end
+RUN humann_test --run-functional-tests-end-to-end
 
-#RUN cd humann-3.9 && ls examples &&  humann --input examples/demo.fasta.gz --output tmpout
+RUN cd humann-3.9 && ls examples &&  humann --input examples/demo.fasta.gz --output tmpout
diff --git a/biobakery-profiler/config_4.0.0.alpha.1-final-smaller-v2.patch b/biobakery-profiler/config_4.0.0.alpha.1-final-smaller-v2-with_tests.patch
similarity index 100%
rename from biobakery-profiler/config_4.0.0.alpha.1-final-smaller-v2.patch
rename to biobakery-profiler/config_4.0.0.alpha.1-final-smaller-v2-with_tests.patch
diff --git a/biobakery-profiler/nucleotide_4.0.0.alpha.1-final-smaller-v2.patch b/biobakery-profiler/nucleotide_4.0.0.alpha.1-final-smaller-v2-with_tests.patch
similarity index 100%
rename from biobakery-profiler/nucleotide_4.0.0.alpha.1-final-smaller-v2.patch
rename to biobakery-profiler/nucleotide_4.0.0.alpha.1-final-smaller-v2-with_tests.patch
diff --git a/biobakery_build_manifest.csv b/biobakery_build_manifest.csv
index 9cc97e4..065d9a5 100644
--- a/biobakery_build_manifest.csv
+++ b/biobakery_build_manifest.csv
@@ -1,3 +1,3 @@
-biobakery-profiler,4.0.6--4.0.0.alpha.1-final-smaller-v2
+biobakery-profiler,4.0.6--4.0.0.alpha.1-final-smaller-v2-with_tests
 biobakery-profiler,4.1.0--v3.9
 biobakery-profiler,4.0.5--3.6.1

From b7b2cfd2b6de13899014a2a5d558a4b3c2716dd2 Mon Sep 17 00:00:00 2001
From: miraep8 <miraep8@gmail.com>
Date: Fri, 10 Apr 2026 13:06:37 -0400
Subject: [PATCH 08/17] fix image names

---
 ...4.0.0.alpha.1-final-smaller-v2-with_tests.patch | 14 --------------
 ....patch => nucleotide_4.0.0.alpha.1-final.patch} |  0
 biobakery_build_manifest.csv                       |  2 +-
 3 files changed, 1 insertion(+), 15 deletions(-)
 delete mode 100644 biobakery-profiler/config_4.0.0.alpha.1-final-smaller-v2-with_tests.patch
 rename biobakery-profiler/{nucleotide_4.0.0.alpha.1-final-smaller-v2-with_tests.patch => nucleotide_4.0.0.alpha.1-final.patch} (100%)

diff --git a/biobakery-profiler/config_4.0.0.alpha.1-final-smaller-v2-with_tests.patch b/biobakery-profiler/config_4.0.0.alpha.1-final-smaller-v2-with_tests.patch
deleted file mode 100644
index c5246b8..0000000
--- a/biobakery-profiler/config_4.0.0.alpha.1-final-smaller-v2-with_tests.patch
+++ /dev/null
@@ -1,14 +0,0 @@
-105c105,107
-< user_edit_config_file="humann.cfg"
----
-> # User config file
-> if not os.environ.get("HUMANN_CONFIG"):
->     user_edit_config_file="humann.cfg"
-107,108c109,112
-< full_path_user_edit_config_file=os.path.join(os.path.dirname(os.path.abspath(__file__)),
-<     user_edit_config_file)
----
->     full_path_user_edit_config_file=os.path.join(os.path.dirname(os.path.abspath(__file__)),
->                                                  user_edit_config_file)
-> else:
->     full_path_user_edit_config_file=os.environ.get("HUMANN_CONFIG")
diff --git a/biobakery-profiler/nucleotide_4.0.0.alpha.1-final-smaller-v2-with_tests.patch b/biobakery-profiler/nucleotide_4.0.0.alpha.1-final.patch
similarity index 100%
rename from biobakery-profiler/nucleotide_4.0.0.alpha.1-final-smaller-v2-with_tests.patch
rename to biobakery-profiler/nucleotide_4.0.0.alpha.1-final.patch
diff --git a/biobakery_build_manifest.csv b/biobakery_build_manifest.csv
index 065d9a5..8252386 100644
--- a/biobakery_build_manifest.csv
+++ b/biobakery_build_manifest.csv
@@ -1,3 +1,3 @@
-biobakery-profiler,4.0.6--4.0.0.alpha.1-final-smaller-v2-with_tests
+biobakery-profiler,4.0.6_smaller--4.0.0.alpha.1-final
 biobakery-profiler,4.1.0--v3.9
 biobakery-profiler,4.0.5--3.6.1

From e6f04ac9976672b90a61dacb7fc383f199c096bd Mon Sep 17 00:00:00 2001
From: miraep8 <miraep8@gmail.com>
Date: Fri, 10 Apr 2026 13:15:13 -0400
Subject: [PATCH 09/17] comment out tests again

---
 biobakery-profiler/Dockerfile | 12 ++++++------
 1 file changed, 6 insertions(+), 6 deletions(-)

diff --git a/biobakery-profiler/Dockerfile b/biobakery-profiler/Dockerfile
index 592de00..bfdd8f7 100644
--- a/biobakery-profiler/Dockerfile
+++ b/biobakery-profiler/Dockerfile
@@ -148,9 +148,9 @@ ENV LANG=en_US.UTF-8
 WORKDIR /tmp
 
 # The --help is so you can check to ensure that the grouping table are properly read
-RUN humann_config --update database_folders utility_mapping /dbs/util/ && \
-    humann_regroup_table --help
-
-RUN humann_test --run-functional-tests-end-to-end
-
-RUN cd humann-3.9 && ls examples &&  humann --input examples/demo.fasta.gz --output tmpout
+#RUN humann_config --update database_folders utility_mapping /dbs/util/ && \
+#    humann_regroup_table --help
+#
+#RUN humann_test --run-functional-tests-end-to-end
+#
+#RUN cd humann-3.9 && ls examples &&  humann --input examples/demo.fasta.gz --output tmpout

From b4082e49ac87f3fa052cf15d1f713abce53ca4d8 Mon Sep 17 00:00:00 2001
From: miraep8 <miraep8@gmail.com>
Date: Fri, 10 Apr 2026 15:03:09 -0400
Subject: [PATCH 10/17] tries to fix order of install

---
 biobakery-profiler/Dockerfile | 116 ++++++++--------------------------
 biobakery_build_manifest.csv  |   2 +-
 2 files changed, 27 insertions(+), 91 deletions(-)

diff --git a/biobakery-profiler/Dockerfile b/biobakery-profiler/Dockerfile
index bfdd8f7..79cea3f 100644
--- a/biobakery-profiler/Dockerfile
+++ b/biobakery-profiler/Dockerfile
@@ -8,40 +8,28 @@ RUN apt-get update && \
         wget ca-certificates build-essential zlib1g-dev libbz2-dev liblzma-dev && \
     rm -rf /var/lib/apt/lists/*
 
-#old command that was at this stage, commented out:
-# RUN mkdir -p /dbs/util/ && \
-#    wget  -c --read-timeout=5 http://huttenhower.sph.harvard.edu/humann_data/full_mapping_v201901b.tar.gz && \
-#    cd /dbs/util/ && \
-#    tar xzf /full_mapping_v201901b.tar.gz && \
-#    rm /full_mapping_v201901b.tar.gz    
-
 RUN wget -q https://github.com/samtools/samtools/releases/download/1.15.1/samtools-1.15.1.tar.bz2 && \
     tar xjf samtools-1.15.1.tar.bz2 && \
     cd samtools-1.15.1 && \
     ./configure --without-curses && \
     make -j$(nproc) && \
-    make install && \
-    which samtools
+    make install
 
 # ============================================================
 # Stage 2: Build Python packages in a venv we can copy
 # ============================================================
-
 FROM python:3.9.17-slim-bookworm AS pybuild
 
 ARG HUMANNVERSION=0.0.0
 ARG METAPHLANVERSION=0.0.0
 
+# libglpk-dev needed at compile time for Python GLPK bindings
+# see: https://forum.biobakery.org/t/silent-errors-with-bad-install-of-libglpk/4814
 RUN apt-get update && \
     DEBIAN_FRONTEND="noninteractive" apt-get install --no-install-recommends -y \
-        build-essential \
-        wget \
-        ca-certificates \
-        patch \
-        libjpeg-dev \
-        zlib1g-dev \
-        libbz2-dev \
-        liblzma-dev && \
+        build-essential git patch \
+        libjpeg-dev zlib1g-dev libbz2-dev liblzma-dev \
+        libglpk-dev glpk-utils && \
     rm -rf /var/lib/apt/lists/*
 
 RUN python3 -m venv /opt/venv
@@ -50,101 +38,49 @@ ENV PATH="/opt/venv/bin:$PATH"
 COPY nucleotide_${HUMANNVERSION}.patch /tmp/nucleotide.patch
 COPY config_${HUMANNVERSION}.patch /tmp/config.patch
 
-# Install prerequisites
-RUN pip install --no-cache-dir numpy cython && \
-    pip install --no-cache-dir boto3 cloudpickle awscli && \
-    pip install --no-cache-dir biom-format
+# Install order matters: numpy → biom-format → humann → metaphlan
+RUN pip install --no-cache-dir numpy cython
+RUN pip install --no-cache-dir biom-format
+RUN pip install --no-cache-dir boto3 cloudpickle awscli
 
-# Download and inspect humann — STOP HERE to debug
-RUN set -e && \
-    echo "=== HUMANNVERSION is: ${HUMANNVERSION} ===" && \
-    wget -qO /tmp/humann.tar.gz https://github.com/biobakery/humann/archive/refs/tags/${HUMANNVERSION}.tar.gz && \
-    echo "=== Tarball contents (top-level) ===" && \
-    tar tzf /tmp/humann.tar.gz | head -20 && \
-    tar xzf /tmp/humann.tar.gz -C /tmp && \
-    echo "=== Directories in /tmp ===" && \
-    ls -la /tmp/ && \
-    echo "=== Moving to /tmp/humann ===" && \
-    mv /tmp/humann-* /tmp/humann && \
-    ls -la /tmp/humann/ && \
-    echo "=== setup.py exists? ===" && \
-    cat /tmp/humann/setup.py | head -30
-
-# Patch and modify setup.py
 RUN set -e && \
+    git clone --depth=1 --branch ${HUMANNVERSION} https://github.com/biobakery/humann.git /tmp/humann && \
     cd /tmp/humann && \
-    echo "=== Applying patches ===" && \
     patch humann/search/nucleotide.py < /tmp/nucleotide.patch && \
     patch humann/config.py < /tmp/config.patch && \
-    cp setup.py setup.py.orig && \
     sed -i 's|2\.2\.3|2\.5\.1|g' setup.py && \
     sed -i 's|bowtie2_folder="bowtie2-2\.5\.1|bowtie2_folder="bowtie2-2\.5\.1-linux-x86_64|g' setup.py && \
-    echo "=== setup.py diff ===" && \
-    diff setup.py.orig setup.py || true && \
-    echo "=== Full setup.py ===" && \
-    cat setup.py
+    pip install --no-cache-dir . --no-binary :all: && \
+    ldconfig /usr/local/lib && \
+    rm -rf /tmp/humann /tmp/*.patch
+
+RUN pip install --no-cache-dir metaphlan==${METAPHLANVERSION}
+
+# Verify everything works
+RUN python3 -c "from biom import load_table; print('biom OK')" && \
+    humann --version && \
+    metaphlan --version
 
-# Install humann — this is the step we need to watch carefully
-RUN set -e && \
-    echo "=== which pip ===" && \
-    which pip && \
-    echo "=== which python ===" && \
-    which python && \
-    echo "=== pip location ===" && \
-    pip --version && \
-    cd /tmp/humann && \
-    echo "=== Installing humann ===" && \
-    pip install --no-cache-dir . --no-binary :all: 2>&1 && \
-    echo "=== pip install exit code: $? ===" && \
-    echo "=== Checking install ===" && \
-    pip show humann && \
-    echo "=== venv bin contents ===" && \
-    ls -la /opt/venv/bin/ && \
-    echo "=== humann files in venv ===" && \
-    ls /opt/venv/bin/humann* || echo "NO HUMANN BINARIES FOUND" && \
-    echo "=== entry_points check ===" && \
-    pip show -f humann | head -50 && \
-    echo "=== system-wide check ===" && \
-    find / -name "humann*" -type f 2>/dev/null | head -20
 # ============================================================
 # Stage 3: Minimal runtime image
 # ============================================================
 FROM python:3.9.17-slim-bookworm AS runtime-image
 
-ARG HUMANNVERSION=0.0.0
-ARG METAPHLANVERSION=0.0.0
-
-# Copy samtools binary and its needed libs
 COPY --from=dbbuild /usr/local/bin/samtools /usr/local/bin/samtools
-
-# Copy the entire pre-built venv
 COPY --from=pybuild /opt/venv /opt/venv
 ENV PATH="/opt/venv/bin:$PATH"
 
-# You may be tempted to do these all in one call but unfortunately
-# (eg) biom-format wont build without numpy, and it wont install it automatically
-
-# see the glpsol issue here https://forum.biobakery.org/t/silent-errors-with-bad-install-of-libglpk/4814
-# Install ONLY runtime dependencies — no build-essential, no git
+# Runtime libs (not -dev) matching what was compiled against in pybuild
 RUN apt-get update && \
     DEBIAN_FRONTEND="noninteractive" apt-get install --no-install-recommends -y \
-        libjpeg62-turbo \
-        zlib1g \
-        libbz2-1.0 \
-        liblzma5 \
-        ncbi-blast+ \
-        mafft \
-        raxml \
-        libgomp1 \
-        libglpk40 \
-        glpk-utils && \
+        libjpeg62-turbo zlib1g libbz2-1.0 liblzma5 \
+        libglpk40 glpk-utils libgomp1 \
+        ncbi-blast+ mafft raxml && \
     rm -rf /var/lib/apt/lists/* && \
-    ldconfig /usr/local/lib && \
-    glpsol --version
+    ldconfig /usr/local/lib
 
 ENV LC_ALL=en_US.UTF-8
 ENV LANG=en_US.UTF-8
-
 WORKDIR /tmp
 
 # The --help is so you can check to ensure that the grouping table are properly read
diff --git a/biobakery_build_manifest.csv b/biobakery_build_manifest.csv
index 8252386..dcbeb21 100644
--- a/biobakery_build_manifest.csv
+++ b/biobakery_build_manifest.csv
@@ -1,3 +1,3 @@
-biobakery-profiler,4.0.6_smaller--4.0.0.alpha.1-final
+biobakery-profiler,4.0.7_smaller--4.0.0.alpha.1-final
 biobakery-profiler,4.1.0--v3.9
 biobakery-profiler,4.0.5--3.6.1

From 3d791fcc75131af75b541d9c05c3a2df693ea6eb Mon Sep 17 00:00:00 2001
From: miraep8 <miraep8@gmail.com>
Date: Fri, 10 Apr 2026 15:16:58 -0400
Subject: [PATCH 11/17] try again full not simplified dockerfile with tests

---
 biobakery-profiler/Dockerfile | 104 ++++++++++++++++++++++++++--------
 1 file changed, 81 insertions(+), 23 deletions(-)

diff --git a/biobakery-profiler/Dockerfile b/biobakery-profiler/Dockerfile
index 79cea3f..8d893f9 100644
--- a/biobakery-profiler/Dockerfile
+++ b/biobakery-profiler/Dockerfile
@@ -13,7 +13,8 @@ RUN wget -q https://github.com/samtools/samtools/releases/download/1.15.1/samtoo
     cd samtools-1.15.1 && \
     ./configure --without-curses && \
     make -j$(nproc) && \
-    make install
+    make install && \
+    which samtools
 
 # ============================================================
 # Stage 2: Build Python packages in a venv we can copy
@@ -23,13 +24,19 @@ FROM python:3.9.17-slim-bookworm AS pybuild
 ARG HUMANNVERSION=0.0.0
 ARG METAPHLANVERSION=0.0.0
 
-# libglpk-dev needed at compile time for Python GLPK bindings
-# see: https://forum.biobakery.org/t/silent-errors-with-bad-install-of-libglpk/4814
+# IMPORTANT: libglpk-dev is needed here at compile time
+# so that the Python GLPK bindings compile correctly
 RUN apt-get update && \
     DEBIAN_FRONTEND="noninteractive" apt-get install --no-install-recommends -y \
-        build-essential git patch \
-        libjpeg-dev zlib1g-dev libbz2-dev liblzma-dev \
-        libglpk-dev glpk-utils && \
+        build-essential \
+        git \
+        patch \
+        libjpeg-dev \
+        zlib1g-dev \
+        libbz2-dev \
+        liblzma-dev \
+        libglpk-dev \
+        glpk-utils && \
     rm -rf /var/lib/apt/lists/*
 
 RUN python3 -m venv /opt/venv
@@ -38,28 +45,58 @@ ENV PATH="/opt/venv/bin:$PATH"
 COPY nucleotide_${HUMANNVERSION}.patch /tmp/nucleotide.patch
 COPY config_${HUMANNVERSION}.patch /tmp/config.patch
 
-# Install order matters: numpy → biom-format → humann → metaphlan
+# IMPORTANT: Installation order matters!
+# 1. numpy and cython MUST be installed before biom-format
+# 2. biom-format must be installed before humann
+# 3. GLPK libs must be present when installing packages that link against them
+# see: https://forum.biobakery.org/t/silent-errors-with-bad-install-of-libglpk/4814
+
+# Step 1: numpy and cython first — biom-format needs these at build time
 RUN pip install --no-cache-dir numpy cython
-RUN pip install --no-cache-dir biom-format
+
+# Step 2: biom-format — needs numpy already installed to compile
+RUN pip install --no-cache-dir biom-format && \
+    python3 -c "from biom import load_table; print('biom-format OK')"
+
+# Step 3: Other Python dependencies
 RUN pip install --no-cache-dir boto3 cloudpickle awscli
 
+# Step 4: humann — needs biom-format already working
 RUN set -e && \
     git clone --depth=1 --branch ${HUMANNVERSION} https://github.com/biobakery/humann.git /tmp/humann && \
     cd /tmp/humann && \
     patch humann/search/nucleotide.py < /tmp/nucleotide.patch && \
     patch humann/config.py < /tmp/config.patch && \
+    cp setup.py setup.py.orig && \
     sed -i 's|2\.2\.3|2\.5\.1|g' setup.py && \
     sed -i 's|bowtie2_folder="bowtie2-2\.5\.1|bowtie2_folder="bowtie2-2\.5\.1-linux-x86_64|g' setup.py && \
+    diff setup.py.orig setup.py || true && \
     pip install --no-cache-dir . --no-binary :all: && \
     ldconfig /usr/local/lib && \
-    rm -rf /tmp/humann /tmp/*.patch
+    glpsol --version
 
+# Step 5: metaphlan after humann
 RUN pip install --no-cache-dir metaphlan==${METAPHLANVERSION}
 
-# Verify everything works
-RUN python3 -c "from biom import load_table; print('biom OK')" && \
+# Verify the full dependency chain works
+RUN set -e && \
+    echo "=== Verifying dependency chain ===" && \
+    python3 -c "import numpy; print(f'numpy {numpy.__version__} OK')" && \
+    python3 -c "from biom import load_table; print('biom-format OK')" && \
+    python3 -c "import humann; print('humann OK')" && \
+    python3 -c "import metaphlan; print('metaphlan OK')" && \
+    which humann && \
     humann --version && \
-    metaphlan --version
+    ls /opt/venv/bin/humann* && \
+    echo "=== All pybuild checks passed ==="
+
+# Clean up
+RUN find /opt/venv -type d -name "__pycache__" -exec rm -rf {} + 2>/dev/null; \
+    find /opt/venv -type d -name "tests" -exec rm -rf {} + 2>/dev/null; \
+    find /opt/venv -type d -name "test" -exec rm -rf {} + 2>/dev/null; \
+    find /opt/venv -name "*.pyc" -delete 2>/dev/null; \
+    find /opt/venv -name "*.pyo" -delete 2>/dev/null; \
+    rm -rf /tmp/humann /tmp/*.patch
 
 # ============================================================
 # Stage 3: Minimal runtime image
@@ -70,12 +107,20 @@ COPY --from=dbbuild /usr/local/bin/samtools /usr/local/bin/samtools
 COPY --from=pybuild /opt/venv /opt/venv
 ENV PATH="/opt/venv/bin:$PATH"
 
-# Runtime libs (not -dev) matching what was compiled against in pybuild
+# IMPORTANT: Runtime versions of the same libraries that were used at compile time
+# libglpk40 + glpk-utils must be here or humann/biom will fail at runtime
 RUN apt-get update && \
     DEBIAN_FRONTEND="noninteractive" apt-get install --no-install-recommends -y \
-        libjpeg62-turbo zlib1g libbz2-1.0 liblzma5 \
-        libglpk40 glpk-utils libgomp1 \
-        ncbi-blast+ mafft raxml && \
+        libjpeg62-turbo \
+        zlib1g \
+        libbz2-1.0 \
+        liblzma5 \
+        ncbi-blast+ \
+        mafft \
+        raxml \
+        libgomp1 \
+        libglpk40 \
+        glpk-utils && \
     rm -rf /var/lib/apt/lists/* && \
     ldconfig /usr/local/lib
 
@@ -83,10 +128,23 @@ ENV LC_ALL=en_US.UTF-8
 ENV LANG=en_US.UTF-8
 WORKDIR /tmp
 
-# The --help is so you can check to ensure that the grouping table are properly read
-#RUN humann_config --update database_folders utility_mapping /dbs/util/ && \
-#    humann_regroup_table --help
-#
-#RUN humann_test --run-functional-tests-end-to-end
-#
-#RUN cd humann-3.9 && ls examples &&  humann --input examples/demo.fasta.gz --output tmpout
+# Thorough smoke tests — verify the full chain works at runtime
+RUN set -e && \
+    echo "=== Runtime smoke tests ===" && \
+    echo "--- Binary checks ---" && \
+    glpsol --version && \
+    samtools --version | head -1 && \
+    echo "--- Python import chain (order matters) ---" && \
+    python3 -c "import numpy; print(f'numpy {numpy.__version__} OK')" && \
+    python3 -c "from biom import load_table; print('biom-format OK')" && \
+    python3 -c "import humann; print('humann OK')" && \
+    python3 -c "import metaphlan; print('metaphlan OK')" && \
+    echo "--- HUMAnN tool checks ---" && \
+    humann --version && \
+    humann_config --print && \
+    humann_regroup_table --help > /dev/null && \
+    humann_renorm_table --help > /dev/null && \
+    humann_join_tables --help > /dev/null && \
+    echo "--- MetaPhlAn check ---" && \
+    metaphlan --version && \
+    echo "=== All runtime smoke tests passed ==="
\ No newline at end of file

From 53b4c976e2150f7dad75791ba6c6d620ff4dbdd3 Mon Sep 17 00:00:00 2001
From: miraep8 <miraep8@gmail.com>
Date: Fri, 10 Apr 2026 16:04:07 -0400
Subject: [PATCH 12/17] fix parsing of version

---
 .github/workflows/bb_builder.yaml | 2 +-
 biobakery_build_manifest.csv      | 2 +-
 2 files changed, 2 insertions(+), 2 deletions(-)

diff --git a/.github/workflows/bb_builder.yaml b/.github/workflows/bb_builder.yaml
index c34feea..465f14c 100644
--- a/.github/workflows/bb_builder.yaml
+++ b/.github/workflows/bb_builder.yaml
@@ -45,7 +45,7 @@ jobs:
         env:
           tool_ver: ${{ matrix.manifest }}
         id: humannversplit
-        run: TMP=${tool_ver##*,} && echo "HUMANNVERSION=${TMP##*--}" >> $GITHUB_OUTPUT
+        run: TMP=${tool_ver##*,} && echo "AFTER_DASH=${TMP##*--}" && echo "HUMANNVERSION=${AFTER_DASH%%;*}" >> $GITHUB_OUTPUT
       - name: Get metaphlan version
         env:
           tool_ver: ${{ matrix.manifest }}
diff --git a/biobakery_build_manifest.csv b/biobakery_build_manifest.csv
index dcbeb21..5061feb 100644
--- a/biobakery_build_manifest.csv
+++ b/biobakery_build_manifest.csv
@@ -1,3 +1,3 @@
-biobakery-profiler,4.0.7_smaller--4.0.0.alpha.1-final
+biobakery-profiler,4.0.6--4.0.0.alpha.1-final;smaller
 biobakery-profiler,4.1.0--v3.9
 biobakery-profiler,4.0.5--3.6.1

From 436348478615d8f56385be0d12755c0fc4fae7fb Mon Sep 17 00:00:00 2001
From: miraep8 <miraep8@gmail.com>
Date: Fri, 10 Apr 2026 16:07:19 -0400
Subject: [PATCH 13/17] better symbol breaking

---
 .github/workflows/bb_builder.yaml | 2 +-
 biobakery_build_manifest.csv      | 2 +-
 2 files changed, 2 insertions(+), 2 deletions(-)

diff --git a/.github/workflows/bb_builder.yaml b/.github/workflows/bb_builder.yaml
index 465f14c..e2330c0 100644
--- a/.github/workflows/bb_builder.yaml
+++ b/.github/workflows/bb_builder.yaml
@@ -45,7 +45,7 @@ jobs:
         env:
           tool_ver: ${{ matrix.manifest }}
         id: humannversplit
-        run: TMP=${tool_ver##*,} && echo "AFTER_DASH=${TMP##*--}" && echo "HUMANNVERSION=${AFTER_DASH%%;*}" >> $GITHUB_OUTPUT
+        run: TMP=${tool_ver##*,} && echo "AFTER_DASH=${TMP##*--}" && echo "HUMANNVERSION=${AFTER_DASH%%_*}" >> $GITHUB_OUTPUT
       - name: Get metaphlan version
         env:
           tool_ver: ${{ matrix.manifest }}
diff --git a/biobakery_build_manifest.csv b/biobakery_build_manifest.csv
index 5061feb..6f035b2 100644
--- a/biobakery_build_manifest.csv
+++ b/biobakery_build_manifest.csv
@@ -1,3 +1,3 @@
-biobakery-profiler,4.0.6--4.0.0.alpha.1-final;smaller
+biobakery-profiler,4.0.6--4.0.0.alpha.1-final_smaller
 biobakery-profiler,4.1.0--v3.9
 biobakery-profiler,4.0.5--3.6.1

From 57def9bb66a9accc20d89627858ce342fb139a41 Mon Sep 17 00:00:00 2001
From: miraep8 <miraep8@gmail.com>
Date: Fri, 10 Apr 2026 16:12:14 -0400
Subject: [PATCH 14/17] actually set middleman variable

---
 .github/workflows/bb_builder.yaml | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/.github/workflows/bb_builder.yaml b/.github/workflows/bb_builder.yaml
index e2330c0..65916a3 100644
--- a/.github/workflows/bb_builder.yaml
+++ b/.github/workflows/bb_builder.yaml
@@ -45,7 +45,7 @@ jobs:
         env:
           tool_ver: ${{ matrix.manifest }}
         id: humannversplit
-        run: TMP=${tool_ver##*,} && echo "AFTER_DASH=${TMP##*--}" && echo "HUMANNVERSION=${AFTER_DASH%%_*}" >> $GITHUB_OUTPUT
+        run: TMP=${tool_ver##*,} && AFTER_DASH=${TMP#*--} && echo "HUMANNVERSION=${AFTER_DASH%%_*}" >> $GITHUB_OUTPUT
       - name: Get metaphlan version
         env:
           tool_ver: ${{ matrix.manifest }}

From 72dc5e9c1ab47610fe45d84913547e947716fcc0 Mon Sep 17 00:00:00 2001
From: miraep8 <miraep8@gmail.com>
Date: Fri, 10 Apr 2026 17:24:07 -0400
Subject: [PATCH 15/17] add things nextflow needs

---
 biobakery-profiler/Dockerfile | 17 ++++++++++++++++-
 1 file changed, 16 insertions(+), 1 deletion(-)

diff --git a/biobakery-profiler/Dockerfile b/biobakery-profiler/Dockerfile
index 8d893f9..0d574c3 100644
--- a/biobakery-profiler/Dockerfile
+++ b/biobakery-profiler/Dockerfile
@@ -147,4 +147,19 @@ RUN set -e && \
     humann_join_tables --help > /dev/null && \
     echo "--- MetaPhlAn check ---" && \
     metaphlan --version && \
-    echo "=== All runtime smoke tests passed ==="
\ No newline at end of file
+    echo "=== All runtime smoke tests passed ==="
+
+RUN apt-get update && \
+    DEBIAN_FRONTEND="noninteractive" apt-get install --no-install-recommends -y \
+        libjpeg62-turbo zlib1g libbz2-1.0 liblzma5 \
+        libglpk40 glpk-utils libgomp1 \
+        ncbi-blast+ mafft raxml \
+        procps \
+        locales && \
+    rm -rf /var/lib/apt/lists/* && \
+    ldconfig /usr/local/lib && \
+    sed -i '/en_US.UTF-8/s/^# //g' /etc/locale.gen && \
+    locale-gen en_US.UTF-8
+
+ENV LC_ALL=en_US.UTF-8
+ENV LANG=en_US.UTF-8
\ No newline at end of file

From 553e442282a753ee93145dd2020a1a2481d3cc17 Mon Sep 17 00:00:00 2001
From: miraep8 <miraep8@gmail.com>
Date: Fri, 10 Apr 2026 17:25:11 -0400
Subject: [PATCH 16/17] update image stuff

---
 biobakery_build_manifest.csv | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/biobakery_build_manifest.csv b/biobakery_build_manifest.csv
index 6f035b2..871342b 100644
--- a/biobakery_build_manifest.csv
+++ b/biobakery_build_manifest.csv
@@ -1,3 +1,3 @@
-biobakery-profiler,4.0.6--4.0.0.alpha.1-final_smaller
+biobakery-profiler,4.0.6--4.0.0.alpha.1-final_smaller-pt2
 biobakery-profiler,4.1.0--v3.9
 biobakery-profiler,4.0.5--3.6.1

From e645c7d55964d52c6e88bffc60a14f40dbbce5cb Mon Sep 17 00:00:00 2001
From: miraep8 <miraep8@gmail.com>
Date: Fri, 10 Apr 2026 19:02:48 -0400
Subject: [PATCH 17/17] make humann3 smaller as well

---
 biobakery_build_manifest.csv | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/biobakery_build_manifest.csv b/biobakery_build_manifest.csv
index 871342b..8bc8841 100644
--- a/biobakery_build_manifest.csv
+++ b/biobakery_build_manifest.csv
@@ -1,3 +1,3 @@
 biobakery-profiler,4.0.6--4.0.0.alpha.1-final_smaller-pt2
 biobakery-profiler,4.1.0--v3.9
-biobakery-profiler,4.0.5--3.6.1
+biobakery-profiler,4.0.5--3.6.1_smaller-pt2