Command Line Options

Command line options can be listed by running fastproject with the -h argument:

fastproject -h

Full List

• data_file

Input expression matrix file. This input is required and should go first. For formatting instructions see the Input file formats page

• -k FILE or --housekeeping FILE

Optional Text file specifying houskeeping gene list to use when estimating dropout probabilities. One gene per line, identifiers should match row labels of data_file

If the list isn't specified, a default list is used

• -s [FILE [FILE ...]] or --signatures [FILE [FILE ...]]

Files containing gene signatures. Many files can be supplied here separated by spaces

• -p [FILE [FILE ...]] or --precomputed [FILE [FILE ...]]

Files containing meta-data about the cells/samples. Labels (denoting information such as chemical stimuli) or numerical information (such as exposure time) are acceptable. For formatting instructions see the Input file formats page

*Note: -s and -p are not both required, but one of these options must be selected

• -o DIRECTORY or --output DIRECTORY

Name of output directory. Otherwise, output directory is auto-generated

• --nofilter

Use all genes when creating projections. Useful if data has been pre-filtered

• --nomodel

No estimation of expression probability or false negative probability Useful when input matrix is not single-cell data (population RNA-seq or lipidomics data, for example)

• --pca_filter

Filters PC principal components that correlate with a QC metric (area under the false-negative curve) for each sample

• --qc

Performs a quality check on samples, filtering samples that do not pass

• --all_sigs

Do not remove insignificant signatures from output. Normally, FastProject limits the output HTML report to only include signatures that were significantly associated with at least one projection metric. This is done in order to decrease output report size and report loading time.

• --debug

Run FastProject in Debug mode Useful for developing on FastProject. Python's pdb debugger is invoked on error.

• --min_signature_genes N

Signatures that match less than N genes in the data are discarded This is optional. If not specified, the default threshold is 5 genes

• --projections [FILE [FILE ...]]

Loads projection coordinates from FILE. Useful if you want to evaluate signatures against a projection generated with another tool. For formatting instructions see the Input file formats page

• --weights FILE

Loads weights from FILE. Use these weights instead of FastProject's FNR calculation

• --threshold N

Removes transcripts detected in less than N samples. If not specified, default is 20% of total sample count.

• --sig_norm_method {none,znorm_columns,znorm_rows,znorm_rows_then_columns,rank_norm_columns}

Pick a normalization method to be applied to data before evaluating signature scores If not specified, default is z-normalize the rows of the expression matrix.

• --sig_score_method {naive,weighted_avg,imputed,only_nonzero}

Pick a method to evaluate signature scores If not specified, default is the weighted-avg method using FN probabilities as weights.