Home

Welcome to the GeneParliamentID wiki! This page offers a broad overview of the GeneParliamentID (GPID) pipeline.

For details on the following topics, please visit the dedicated wiki pages:

• Setup
• Pipeline parameters
• Method calibration
• Interpretation
• Tutorial

Inputs

• A set of multipe genes for a sample of unknown identity
• A reference data set containing the same genes for the lineage of interest
• Calibration files to set pipeline parameters
• Optional: Classification file that specifies user-defined groups of closely related species

For detailed instructions, see Pipeline parameters.

Key pipeline steps

1. Select genes with performance of species identification above user-defined threshold
2. Align each gene against reference, select top match based on highest Bit-score
3. Remove low-confidence matches that don't meet the user-defined alignment filtering thresholds
4. Summarise all remaining identifications to the Gene Parliament
5. Flag the sample as data-deficient if the number of genes (parliament size) is below the user-defined threshold
6. Select identification with most support as the top identification
7. Evaluate confidence in the top identification based on the percentage of genes supporting this identification

Outputs

The GeneParliamentID pipeline summarises all individual gene identifications in a Gene Parliament, which represents the percentage of genes supporting all competing identifications. The Gene Parliament is presented both as a table <Sample>_gpid.csv and as a figure <Sample>_gpid.pdf.

For interpretation of these outputs, see Interpretation.

Method calibration

To identify the optimal pipeline parameters for a given lineage, we strongly recommend conducting method calibration using the provided script. Method calibration also requires a set of test samples of known identity. The calibration script helps you to select the optimal parameters that balance accuracy of identification against retrievability of samples.

Method calibration includes the following steps:

1. Select the optimal alignment filtering thresholds
2. Calculate gene performance (percentage of correct identifications) following application of optimal alignment filters
3. Select optimal gene performance threshold
4. Select minimum parliament size threshold
5. Estimate confidence of identification depending on the percentage of genes supporting the top identification