v1.4-beta.2

v1.4-beta.2 is an update to v1.4-beta.1 with bug fixes for variable assignments and random string assignments

SEACR 1.4 development release contains the following changes:

• Added dashed command line parameter support: inputs 1, 2, 3, 4, and 5 now correspond to -b, -c, -n, -m, and -o, respectively
• Added a function to test a window of control scaling thresholds in order to render scaling more stable in "norm" mode
• Added an optional input for a numeric peak extension factor (-e) during peak merging, where all peaks will be extended by e*(mean(peak length)) (default = 0.1)
• Added an optional yes/no input to dictate whether peaks overlapping robust IgG peaks get filtered out or not (-r) (default = "yes")
• Fixed a bug in which empty peak files failed due to a division by zero error before completion
• Fixed a bug in which "NaN" is erroneously passed to quantile() in files with a low complexity of peaks

v1.4-beta.1

SEACR 1.4 development release contains the following changes:

• Added dashed command line parameter support: inputs 1, 2, 3, 4, and 5 now correspond to -b, -c, -n, -m, and -o, respectively
• Added a function to test a window of control scaling thresholds in order to render scaling more stable in "norm" mode
• Added an optional input for a numeric peak extension factor (-e) during peak merging, where all peaks will be extended by e*(mean(peak length)) (default = 0.1)
• Added an optional yes/no input to dictate whether peaks overlapping robust IgG peaks get filtered out or not (-r) (default = "yes")
• Fixed a bug in which empty peak files failed due to a division by zero error before completion
• Fixed a bug in which "NaN" is erroneously passed to quantile() in files with a low complexity of peaks

v1.3

SEACR v1.3 contains the following changes:

• Fixed a bug in which the bedgraph line thresholding added in v1.2 was failing for some datasets.
• Added a check to filter out any input bedgraph lines containing zero signal.

v1.2

SEACR v1.2 contains the following changes:

• Fixed a bug in lines 166 and 168 in which misplaced brackets caused the misreporting of the max signal region terminal coordinate for merged signal blocks
• Added a counter to keep track of the number of component bedgraph lines that compose each signal block, and a function to calculate the minimum threshold of lines per signal block at which there is a smaller percentage of target signal blocks remaining than control. This is meant to be used as a filter for signal blocks that pass the total signal threshold despite being composed of very few bedgraph lines, which are unlikely to be true peaks.
• Changed how the dataframe for density plotting is truncated (previously a hard-coded 90% cutoff): a dataframe of list quantile (i.e. line #/max line#) vs. value quantile (i.e. value/max value) is derived, and the threshold is selected by finding the dataframe pair for which the orthogonal distance below the line defined by (0,0);(1,1) is maximized.

SEACR v1.1

SEACR v1.1 contains the following changes:

• Changed "union" and "AUC" modes to "relaxed" and "stringent" modes, respectively.
• Removed maximum signal threshold from "relaxed" mode and replaced it with an alternate total signal threshold that uses the point halfway between the knee and the peak of the total signal curve as described in the manuscript text. This change improves performance at high read depth.
• Implemented alternate threshold test that searches for any thresholds that come within 95% of the optimal threshold. This change avoids spurious thresholds that are overselective in some datasets.