example_config_file.txt

# barseq_counter configuration file

###### File/folder locations

# Where to put the results!
output_folder: /path/to/your/screen/output


# Sample table (maps index tags and sequence lanes to sample information)
sample_table_file: /path/to/your/screen/sample_table.txt


# Lane location table (maps the lane column in the sample table to the folder
# containing the raw sequencing data for that lane)
lane_location_file: /path/to/your/screen/lane_locations.txt


# Species (required for parsing PCR products, interpreting barcodes,
# and downstream functional analyses)
# Species must correspond to a species name in the species_config file
species_ID: S_cerevisiae_minipool

# File that specifies which barcodes to remove from a dataset. Strains with this barcode survive
# the "per-lane" rounds of processing and are removed in the step that creates the final filtered
# count matrix from all lanes. This is just a '\n'-delimited file of the barcode sequences to be
# removed. This can be left blank if no strains need to be removed.
barcodes_to_remove_file: /path/to/your/screen/remove_some_strains.txt

# Should conditions matching the "barcode-specific" pattern (positive interactions for barcodes
# beginning with one particular base) be removed? (can apply to small or large screens)
# (True / False)
remove_barcode_specific_conditions: True

# Should conditions associated with highly-self-correlating index tags (in control conditions) be removed?
# (Intended for use in larger screens, e.g. 10+ lanes)
# (True / False)
remove_correlated_index_tags: True


###### Miscellaneous parameters
# How verbose should the messages printed to stdout be? Values are:
# 0: no messages
# 1: "announcements," aka "generating z-score matrix for lane X" (default)
# 2: inspection of objects
# 3: deeper inspection of objects, very cryptic
verbosity: 1


##### ADVANCED OPTIONS
##### DO NOT CHANGE UNLESS YOU KNOW WHAT YOU ARE DOING :-)
# Set the number of errors allowed in the common primer and
# genetic barcode (up to and including)
common_primer_tolerance: 2
barcode_tolerance: 2

# Set the read count at which strains in control conditions are "not detected"
# These are set to NA and excluded from the mean control profile
control_detection_limit: 20

# Set the read count at which strains in non-control conditions are "not detected"
# This also applies to control conditions when their normalized profiles are being computed
# (just not when they are used for the reference profile)
# All counts below this number are set to this number. This prevents NAs when taking
# the log of the read counts
sample_detection_limit: 20

# Strain quality cutoff. For a strain to pass, the fraction of conditions with 
# read counts >= "strain_pass_read_count" must be >= "strain_pass_fraction".
strain_pass_read_count: 20
strain_pass_fraction: 0.25

# Condition quality cutoff. For a condition to pass, the fraction of conditions with 
# read counts >= "condition_pass_read_count" must be >= "condition_pass_fraction".
condition_pass_read_count: 20
condition_pass_fraction: 0.25

# Index tag correlation cutoff. If the mean correlation of all DMSO conditions with a particular index tag is
# above "index_tag_correlation_cutoff", all conditions with that index tag are removed from the dataset
index_tag_correlation_cutoff: 0.4

# Barcode-specific correlation to template cutoff. Remove any condition that correlates to any of the
# "barcode-specific" profiles more strongly than this value.
barcode_specific_template_correlation_cutoff: 0.3