ConFindr

What does it do?

Use ConFindr when you think raw sequencing reads may be contaminated or when you need to check raw FASTQ data for contamination before assembly. ConFindr detects evidence of both intra-species and inter-species contamination before assembly or downstream analysis.

Contaminated reads can produce misassemblies and misleading downstream results. ConFindr is therefore a quality-control tool for raw reads, not a general species-identification workflow.

For background, see the ConFindr documentation.

What should I run if I think my raw reads are contaminated?

Run ConFindr when you suspect contamination in raw FASTQ reads. ConFindr checks raw sequencing reads for evidence of intra-species and inter-species contamination before assembly or downstream analysis.

Use ConFindr for a contamination assessment. Do not use Downsample, FastQC, AutoCLARK, or a raw-read retrieval workflow as a substitute for the ConFindr contamination call.

How do I use it?

Subject

In the Subject field, enter:

ConFindr

Spelling matters, but matching is not case-sensitive.

Description

In the Description field, enter one SEQID per line:

2026-SEQ-0001
2026-SEQ-0002

The requested SEQIDs must have raw sequencing reads available.

Attachments

No attachment is required. ConFindr retrieves the raw reads associated with each requested SEQID.

Optional parameters

The supplied documentation does not identify optional parameters for the Redmine ConFindr automator.

Example

See issue 12881 for an example ConFindr request.

Interpreting results

When ConFindr finishes, it uploads:

confindr_output.zip

The archive contains:

confindr_report.csv
confindr_log.txt

confindr_report.csv

The report contains five columns:

  • Sample — the sample identifier;
  • Genus — the genus ConFindr assigns to the sample; more than one genus can be listed when detected;
  • NumContamSNVs — the number of sites that resemble contaminating SNVs;
  • NumUniqueKmers — the number of unique k-mers detected for rMLST genes;
  • ContamStatus — True when ConFindr calls the sample contaminated and False when it calls the sample clean.

The documented contamination criteria are:

  • NumContamSNVs of 3 or greater; or
  • NumUniqueKmers greater than 45000.

Use ContamStatus as the overall workflow call, and review the supporting SNV and k-mer values when investigating a result.

confindr_log.txt

The log is primarily useful for diagnosing unexpected errors. It normally does not need to be interpreted as a biological result.

How long does it take?

ConFindr generally takes approximately one to two minutes per sample. Total runtime depends on read volume, sample count, and service workload.

What can go wrong?

A requested SEQID is unavailable

Symptom: The Redmine issue receives a warning identifying unavailable sequences.

Likely cause: ConFindr cannot locate raw reads for the requested SEQID.

What to do: Verify each SEQID, confirm that its raw sequence files are available, and submit a corrected request.

A contamination call is unexpected

Symptom: ContamStatus is True even though the sample was expected to be clean.

Likely cause: ConFindr detected at least three contamination-like SNVs, more than 45,000 unique rMLST k-mers, or other supporting evidence reflected in the report.

What to do: Review NumContamSNVs, NumUniqueKmers, the assigned genus or genera, raw-read quality, and laboratory context before deciding whether to repeat sequencing or exclude the sample.

The analysis fails unexpectedly

Symptom: No usable report is produced, or the Redmine issue reports an execution error.

Likely cause: The raw reads may be unavailable, malformed, or unsuitable for processing, or the workflow may have encountered a technical error.

What to do: Verify the input files and provide confindr_log.txt to a bioinformatician when escalation is required.

  • AutoCLARK — identifies species represented in raw reads or assemblies; it is not a dedicated contamination call.
  • FastQC/MultiQC — evaluates general raw-read quality but does not make ConFindr contamination calls.
  • Unknown Isolate — identifies an uncertain isolate from a draft genome assembly rather than testing raw reads for contamination.