ampseqQC summarises QC statistics for targeted amplicon sequencing of Plasmodium processed with the Mad4hatter pipeline. It combines the pipeline outputs with a sample manifest to produce an HTML QC report, pass/repool/reprep calls for each sample, control summaries and filtered allele tables.
Documentation: https://eppicenter.github.io/ampseq-qc/, including getting started, an example report and building a custom QC with the package functions.
Installation
# install.packages("remotes")
remotes::install_github("EPPIcenter/ampseq-qc")Rendering the report also requires the Quarto CLI.
Rendering the QC report
library(ampseqQC)
render_qc_report(
results_dir = "path/to/mad4hatter/results",
manifest_file = "path/to/manifest.csv",
output_dir = "qc_output"
)Thresholds can be changed with the arguments of render_qc_report() (see ?render_qc_report), for example read_threshold, reprep_threshold and repool_threshold.
From the command line:
Rscript -e 'ampseqQC::render_qc_report("path/to/results", "path/to/manifest.csv", output_dir = "qc_output")'The report template can be found with qc_report_template() if you want to copy and customise it.
Required inputs
The results directory from the Mad4hatter pipeline must include:
- sample_coverage.txt
- amplicon_coverage.txt
- allele_data.txt
- panel_information/amplicon_info.tsv
If present, allele_data_collapsed.txt and the resistance_marker_module tables are also filtered and written out.
A sample manifest (CSV, TSV or Excel) must contain:
- sample_name – Unique identifier for each sample.
- SampleType – Specifies whether the entry is a sample, positive control, or negative control.
- Batch – Identifies a group of samples processed simultaneously by the same individual.
- Column – The well column where the sample was placed in the plate.
- Row – The well row where the sample was placed in the plate.
- Parasitemia – The qPCR value for the sample.
Panels, pools and reactions
The pools in a run are read from the pool column of panel_information/amplicon_info.tsv, so any combination or subset of pools is supported. Panel settings map each pool to the mPCR reaction it was amplified in, and list targets to exclude from QC calculations (they are still kept in the filtered allele tables).
default_panel_settings() covers the MAD4HatTeR and PfPHAST pools, including versioned and legacy names, using the recommended two-reaction layout:
| Reaction | Pools |
|---|---|
| 1 | D1, D1.1, 1A, R1, R1.1, R1.2, 1B, 5, M1, M1.1, M1.addon |
| 2 | R2, R2.1, 2, M2, M2.1 |
Targets shared by pools in different reactions are counted in each reaction. If a run contains a pool with no reaction defined, the report stops and asks for one. Use panel_settings() to add pools or change the layout:
# Add a bespoke pool
panel <- panel_settings(c(AMPLseq = "3"), base = default_panel_settings())
# R1.2 run in its own reaction
panel <- panel_settings(c(R1 = "3", R1.2 = "3"), base = default_panel_settings())
render_qc_report("path/to/results", "path/to/manifest.csv", panel = panel)Using the functions directly
All of the QC steps are exported, so they can be used outside the report, for example:
manifest <- read_manifest("path/to/manifest.csv")
results <- read_mad4hatter_results("path/to/results", manifest_sample_names = manifest$sample_name)
excluded <- qc_excluded_targets(results$panel_information)
panel_reactions <- assign_reactions(results$panel_information) |>
dplyr::filter(!target_name %in% excluded$excluded)
amplicon_coverage_qc <- results$amplicon_coverage |>
dplyr::filter(!target_name %in% excluded$excluded)
summary_samples <- summarise_samples(
merge_amplicon_coverage(amplicon_coverage_qc, manifest, panel_reactions),
count_targets_per_reaction(panel_reactions),
manifest,
read_threshold = 100
)
qc_calls <- generate_reprep_repool_table(summary_samples) |>
fill_missing_data(manifest)