Add genomes (#45) (#46) #113
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
| # This is a subsampling unit test to get early results | |
| on: | |
| push: | |
| branches: [master, dev, validate-taxonomy] | |
| name: Genera-with-Kraken2 | |
| env: | |
| TSV: "Kalamari/src/genus.tsv" | |
| DB: "kraken2" | |
| SRC_CHR: "Kalamari/src/chromosomes.tsv" | |
| SRC_PLD: "Kalamari/src/plasmids.tsv" | |
| GENUS: Yersinia | |
| jobs: | |
| build: | |
| runs-on: ${{ matrix.os }} | |
| strategy: | |
| matrix: | |
| os: ['ubuntu-20.04' ] | |
| perl: [ '5.32' ] | |
| GENUS: [ 'Yersinia', 'Listeria'] | |
| name: ${{ matrix.GENUS }} Perl ${{ matrix.perl }} on ${{ matrix.os }} | |
| steps: | |
| - name: Set up perl | |
| uses: shogo82148/actions-setup-perl@v1 | |
| with: | |
| perl-version: ${{ matrix.perl }} | |
| multi-thread: "true" | |
| - name: checkout my repo | |
| uses: actions/checkout@v4 | |
| with: | |
| path: Kalamari | |
| - name: env check | |
| run: | | |
| echo $PATH | tr ':' '\n' | sort | |
| - name: apt-get install | |
| run: sudo apt-get install ca-certificates tree jellyfish ncbi-entrez-direct | |
| - name: select for only for this genus | |
| run: | | |
| head -n 1 ${{ env.SRC_CHR }} > ${{ env.TSV }} | |
| grep -m 2 ${{ matrix.GENUS }} ${{ env.SRC_CHR }} >> ${{ env.TSV }} | |
| grep -m 2 ${{ matrix.GENUS }} ${{ env.SRC_PLD }} >> ${{ env.TSV }} | |
| echo "These are the ${{ matrix.GENUS }} genomes for downstream tests" | |
| column -ts $'\t' ${{ env.TSV }} | |
| hexdump -c ${{ env.TSV }} | |
| - name: download | |
| run: | | |
| perl Kalamari/bin/downloadKalamari.pl --outdir ${{ matrix.GENUS }} ${{ env.TSV }} | |
| find ${{ matrix.GENUS }} -name '*.fasta.gz' | xargs gunzip -v | |
| - name: check-results | |
| run: | | |
| tree ${{ matrix.GENUS }} | |
| echo "First two lines of each fasta file:" | |
| find ${{ matrix.GENUS }} -name '*.fasta' | xargs head -n 2 | cut -c 1-60 | |
| - name: install kraken | |
| run: | | |
| wget https://github.com/DerrickWood/kraken2/archive/refs/tags/v2.1.2.tar.gz -O kraken-v2.1.2.tar.gz | |
| tar zxvf kraken-v2.1.2.tar.gz | |
| cd kraken2-2.1.2 && bash install_kraken2.sh target && cd - | |
| ls -lhS kraken2-2.1.2/target | |
| chmod +x kraken2-2.1.2/target/* | |
| - name: build taxonomy | |
| run: | | |
| export PATH=$PATH:Kalamari/bin | |
| buildTaxonomy.sh | |
| ls -lh Kalamari/share | |
| - name: Kraken2 database | |
| run: | | |
| export PATH=$PATH:kraken2-2.1.2/target | |
| which kraken2-build | |
| mkdir -pv ${{ env.DB }} | |
| cp -rv Kalamari/share/kalamari-*/taxonomy ${{ env.DB }}/taxonomy | |
| find ${{ matrix.GENUS }} -name '*.fasta' -exec kraken2-build --db ${{ env.DB }} --add-to-library {} \; | |
| tree ${{ env.DB }} | |
| echo ".....Building the database....." | |
| kraken2-build --build --db ${{ env.DB }} --threads 2 | |
| - name: Kraken2 view results | |
| run: | | |
| export PATH=$PATH:kraken2-2.1.2/target | |
| tree ${{ env.DB }} | |
| ls -lhSR ${{ env.DB }} | |
| QUERY=$(find ${{ matrix.GENUS }} -name '*.fasta' | head -n 1) | |
| echo "QUERY is $QUERY" | |
| set -x; kraken2 --db ${{ env.DB }} --report kraken2.report --use-mpa-style --output kraken2.raw $QUERY; set +x; | |
| head kraken2.report kraken2.raw | |