Add matched DNA and RNA with Docker#

This page adds matched DNA and RNA measurements to a study run in Docker. It follows the first-study tutorial, using its ghcr.io/mannlabs/fasta-lake:local image. De novo selection remains the baseline; matched molecular candidates are added before Sage searches the final FASTA.

Prepare the local inputs#

Use this project layout, with your actual files:

inputs/
  lake.fasta
  samples.tsv
  raw/
    run_A_predictions.csv
    run_A.mzML
    specimen_A_genes.faa
    specimen_A_tpm.tsv
    specimen_A_reference.json

The input specification defines the shared gene IDs, DNA/RNA columns and provider provenance record. Unavailable measurements are missing, not zero. The commands below require matching TPM/protein gene populations; declared incomplete references have a separate coverage policy.

Create destinations and prepare a checked, portable source:

mkdir -p prepared results
docker run --rm --platform linux/amd64 --network none \
  --user "$(id -u):$(id -g)" \
  --mount "type=bind,source=$PWD/inputs,target=/data,readonly" \
  --mount "type=bind,source=$PWD/prepared,target=/prepared" \
  ghcr.io/mannlabs/fasta-lake:local fasta-lake molecular prepare-source \
  --specimen specimen_A --reference-id assembly_A_v1 \
  --tpm /data/raw/specimen_A_tpm.tsv --proteins /data/raw/specimen_A_genes.faa \
  --provenance /data/raw/specimen_A_reference.json --out /prepared/specimen_A

Link the acquisition and check the run#

Save this header and row in inputs/samples.tsv with tab separators:

sample  predictions mzml    molecular_source    specimen
run_A   raw/run_A_predictions.csv   raw/run_A.mzML  ../prepared/specimen_A/source.json  specimen_A

Paths resolve beside the manifest. Repeat the source/specimen pair for additional acquisitions of that specimen; use distinct acquisition IDs.

docker run --rm --platform linux/amd64 --network none \
  --memory 11g --memory-swap 11g --cpus 2 \
  --user "$(id -u):$(id -g)" \
  --mount "type=bind,source=$PWD/inputs,target=/data,readonly" \
  --mount "type=bind,source=$PWD/prepared,target=/prepared,readonly" \
  --mount "type=bind,source=$PWD/results,target=/work" \
  ghcr.io/mannlabs/fasta-lake:local fasta-lake run \
  --manifest /data/samples.tsv --lake /data/lake.fasta --out /work/study_01 \
  --laptop --quantification directlfq \
  --molecular-dna-percent 1 --molecular-rna-percent 1 --validate-only

The two 1% rules are illustrative choices, applied independently to positive DNA and RNA values. Choose and record your rules before evaluating outcomes. Set a layer's percentage to zero to disable it. An 11 GiB container limit does not guarantee that every reference or molecular union fits; consult the resource guide.

Execute and inspect#

Remove --validate-only from the same command to execute. Use a new output name for each run. Inputs and prepared sources stay mounted read-only.

Under results/study_01/:

Output	Meaning
`samples/run_A/molecular/search.fasta`	Preserved de novo targets plus exact additions
`samples/run_A/molecular/gene_evidence.tsv`	Each gene's measurements, selection and sequence identity
`search/run_A/`	Sage results and the actual search configuration
`study_groups/quantification/study_group_matrix.tsv`	directLFQ quantities for the fixed study groups
`study_groups/qc/`	Coverage, missingness and PCA when the study is eligible

One acquisition receives QC but cannot support PCA. The included multi-omics demo provides a tested teaching workflow with real spectra and explicitly synthetic molecular values.