Analyze

Identify differentially abundant genes between the control (the inoculum) and treatment conditions with mbarq analyze

Input/Output Files

Required Inputs

• Count file produced by mbarq merge

barcode	Name	Sample1	Sample2	…
ACCTGGTAG	geneA	500	1000	…
ACCGGGGAA	geneA	100	500	…
CCCGGGAAA	geneB	300	300	…

• Sample data file (CSV) in the following format:

sampleID	treatment
Sample1	control
Sample2	treatment1
…	…

• Name of the column indicating treatment in the sample data should be specified using --treatment_column (for the example above, --treatment_column treatment)

• Treatment level that should be used as a control/baseline should be specified using --baseline (for the example above, --baseline control)

Suggested Inputs

• We highly recommend adding control strains (i.e. strains with barcodes inserted into fitness-neutral locations) to the barcode library. This greatly facilitates quality control and analysis of the data.

• If control strains are present in the library, the control barcodes can be specified with a control file using the --control_file option.

  • In the simplest option, the control file will only contain the barcode sequences of the control strains (1 barcode per line).

  • If different control strains were added at different concentrations, the concentration of each barcode can be specified in the second column.

  • If control strains included strains of different genotypes (ex. wild type as well as negative control strains), the genotype can be specified in the 3rd column.

  • Only wild-type strains will be used for quality control and analysis. This should be specified as wt, WT, or wildtype.

  • The control file should be in CSV format, and contain NO header.

[Required]	[Optional]	[Optional]
ACCTGGGTT	0.005	wt
CCGGAAGGT	0.001	wt

Output Files

• mbarq_merged_counts_batch.txt: Information on sample and batch

• mbarq_merged_counts.correlations.csv: Correlation for each batch

• mbarq_merged_counts_rra_results.csv: Information for each gene about number of barcodes, LFC and false discovery rate

• mbarq_merged_counts_barcodes_results.csv: Information for each barcode about LFC and significance scores

For each comparison:

• mbarq_merged_counts_cond1_vs_cond0.gene_summary.txt: Summary for each gene

• mbarq_merged_counts_cond1_vs_cond0.report.Rmd: MAGeCK Comparison Report

• mbarq_merged_counts_cond1_vs_cond0.sgrna_summary.txt: Summary for each sgRNA

Output Format Options

The final results can be output in two formats:

• Long format (default): Each row represents a gene-treatment combination. This format includes a ‘contrast’ column indicating the treatment condition.

• Wide format: Each row represents a gene, with separate columns for each treatment condition (e.g., ‘LFC_d1’, ‘LFC_d2’). This format is useful for downstream analysis and visualization.

Use the --format option to specify the desired output format.

Format Examples:

Long format (default):

Name	number_of_barcodes	LFC	neg_selection_fdr	pos_selection_fdr	contrast
geneA	3	1.2	0.05	0.9	d1
geneA	3	1.5	0.03	0.8	d2
geneB	2	-0.8	0.8	0.1	d1
geneB	2	-0.9	0.7	0.2	d2

Wide format:

Name	number_of_barcodes	LFC_d1	LFC_d2	neg_selection_fdr_d1	neg_selection_fdr_d2	pos_selection_fdr_d1	pos_selection_fdr_d2
geneA	3	1.2	1.5	0.05	0.03	0.9	0.8
geneB	2	-0.8	-0.9	0.8	0.7	0.1	0.2

Example Usage

# Basic usage with long format output (default)
mbarq analyze -i <count_file> -s <sample_data_file> -c <control_file> \ 
--treatment_column treatment --baseline control 

# Output results in wide format
mbarq analyze -i <count_file> -s <sample_data_file> -c <control_file> \ 
--treatment_column treatment --baseline control --format wide

All Options

mbarq analyze
Usage: mbarq analyze <options>

Options:
  -i, --count_file FILE    CSV file produced by `mbarq merge`
  -s, --sample_data FILE   CSV file containing sample data
  -c, --control_file FILE  control barcode file, see documentation for proper
                           format
  -g, --gene_name STR      column in the count file containing gene
                           identifiers [Name]
  --treatment_column STR   column in sample data file indicating treatment
  --baseline STR           treatment level to use as control/baseline, ex.
                           day0
  -n, --name STR           experiment name, by default will try to use count
                           file name
  -o, --out_dir DIR        Output directory
  --norm_method STR        mageck normalization method: median, total, or 
                           control. By default will use control barcodes if 
                           provided, otherwise median
  --filter_low_counts INT  filter out barcodes with < N reads across all 
                           conditions [0]
  -f, --format STR         output file format: long or wide [long]
  -h, --help               Show this message and exit.