SampleInfo Process Configuration

Purpose

The SampleInfo process is the pipeline entry point that reads sample metadata files, performs statistical analyses, and generates visualization reports.

When to Use

• Always required as first process unless using LoadingRNAFromSeurat
• When you have sample metadata in CSV/TSV format
• When you need to generate statistical summaries and visualizations
• When you want to add or transform metadata columns before downstream analysis

Note: Mutually exclusive with LoadingRNAFromSeurat.

Configuration Structure

[SampleInfo]
cache = true

[SampleInfo.in]
infile = "path/to/sample_info.txt"  # Required: yes (unless using LoadingRNAFromSeurat)

[SampleInfo.envs]
sep = "\t"                  # str - File separator
mutaters = {}                # dict - Column transformations using R expressions
save_mutated = false         # bool - Save mutated columns to output
exclude_cols = "TCRData,BCRData,RNAData"  # Columns hidden in report
defaults = { plot_type = "bar", more_formats = [], save_code = false }
stats = {}                   # dict - Statistical plot definitions

Required input columns: Sample (unique ID), RNAData (directory path) Optional columns: TCRData, BCRData, additional metadata Data format: CSV/TSV with header, RNA data must be Read10X()-compatible

Environment Variables

sep (string): Field separator - "\t", ",", ";", or any character

mutaters (dict): R expressions for dplyr::mutate(). Keys are column names, values are R expressions.

• Example: mutaters = { "AgeGroup" = "ifelse(Age > 60, 'Senior', 'Adult')" }
• Special function paired() identifies paired samples: paired(., 'PatientID', 'Timepoint', c('T1', 'T2'))

save_mutated (bool): Save mutated columns to output file. Factor columns lose level ordering when saved as text.

exclude_cols (str/list): Comma-separated string or list of columns to exclude from report table.

defaults (dict): Default plot parameters inherited by all plots:

[SampleInfo.envs.defaults]
plot_type = "bar"           # Plot type (see External References)
more_formats = []            # Additional formats: ["pdf", "svg"]
save_code = false            # Save R code and data
subset = null                # dplyr::filter expression
section = null               # Report section name
descr = null                 # Plot description
width = null, height = null, res = 100  # Plot dimensions

stats (dict): Plot definitions. Keys are case names (titles), values inherit from defaults.

External References

Plotthis Functions

| plot_type | Function | Description | |-----------|----------|-------------| | pie | PieChart() | Pie chart | | bar | BarPlot() | Bar plot | | box | BoxPlot() | Box plot | | violin | ViolinPlot() | Violin plot | | histogram | Histogram() | Histogram | | density | DensityPlot() | Density plot | | scatter | ScatterPlot() | Scatter plot | | line | LinePlot() | Line plot | | ridge | RidgePlot() | Ridge plot | | heatmap | Heatmap() | Heatmap |

Full reference: https://pwwang.github.io/plotthis/reference/

Common Plot Parameters

x = "column_name", y = "column_name"  # Axis columns
split_by = "column_name", facet_by = "column_name"  # Split/facet
palette = "Paired", alpha = 1.0  # Color and transparency
title = "Plot Title", nrow = 2, ncol = 3  # Layout
legend.position = "right"  # Legend placement

dplyr::filter() for subset

subset = "Sample == 'A'"
subset = "Age > 60"
subset = "Diagnosis %in% c('Colitis', 'Control')"
subset = "Sex == 'F' & Age > 50"

Configuration Examples

Minimal Configuration

[SampleInfo.in]
infile = "samples.txt"

Basic Statistics

[SampleInfo.in]
infile = "sample_info.txt"

[SampleInfo.envs.stats."Samples_per_Diagnosis"]
plot_type = "bar"
x = "Sample"
split_by = "Diagnosis"

Advanced Configuration

[SampleInfo.in]
infile = "metadata/samples.tsv"

[SampleInfo.envs]
save_mutated = true
mutaters = { "AgeGroup" = "ifelse(Age > 60, 'Senior', 'Adult')" }

[SampleInfo.envs.stats."N_Samples_per_Diagnosis"]
x = "Sample"
split_by = "Diagnosis"

[SampleInfo.envs.stats."Age_distribution"]
plot_type = "histogram"
x = "Age"

Common Patterns

Paired Sample Identification

[SampleInfo.envs]
mutaters = { "PairID" = "paired(., 'PatientID', 'Timepoint', c('T1', 'T2'))" }

[SampleInfo.envs.stats."Paired_Samples"]
x = "PairID"
subset = "!is.na(PairID)"

Subset Analysis

[SampleInfo.envs.stats."Controls_Only"]
x = "Sample"
split_by = "Diagnosis"
subset = "Diagnosis == 'Control'"

Dependencies

• Upstream: None (entry point process)
• Downstream: All pipeline processes depend on SampleInfo output
  • SeuratPreparing: Reads sample metadata
  • ScRepLoading: Uses TCRData/BCRData columns
  • All downstream: Use metadata columns for analysis

Validation Rules

Common Errors

1. Missing input file: Always specify [SampleInfo.in.infile]
2. Invalid separator: Match separator to file format (e.g., sep = "," for CSV)
3. Missing required columns: Ensure Sample and RNAData columns exist
4. Factor level ordering: Don't use save_mutated for factor columns - use SeuratPreparing.envs.mutaters instead

Value Constraints

• sep: Single character string
• mutaters: Valid R expressions
• stats keys: Must be unique case names
• devpars.res: Positive integer (default: 100)

Troubleshooting

• Issue: SampleInfo re-runs entire pipeline on parameter change Solution: Set cache = "force" at pipeline level and [SampleInfo] cache = false

• Issue: Factor levels appear in wrong order Solution: Use SeuratPreparing.envs.mutaters for factor columns

• Issue: Plots don't show expected data Solution: Check column names in x, y, split_by match input file exactly

• Issue: Paired sample function returns NA values Solution: Use uniq = false in paired() or adjust idents parameter

• Issue: Mutations not saved for downstream use Solution: Set save_mutated = true. For Seurat metadata, use SeuratPreparing.envs.mutaters

• Issue: Plot type not recognized Solution: Ensure plot_type is lowercase and maps to a plotthis function