ADM3: Automated Detection Algorithm Interpretation.

In ADM3: An Interpretation of the ADM method - automated detection algorithm.

Description

Performs change point detection on a "bed" formatted file.

Usage

ADM3(file, outfile, t=6, np=3, mr=0.3, autoCut=F)

Arguments

file	- "bed" formatted input file (chr, start, stop, value, score).
outfile	- output file location (a report of detected features).
t	- the main detection threshold - controls sensitivity - not recommended below 4.
np	- number of data points required to detect a feature.
mr	- minimum mean ratio to detect a segemnt.
autoCut	- attempt noise dependant auto filtering.

Details

Performs change point detection on a bed formatted file and allows an error value to be applied to each probe.
This is a highly efficent implementation - takes less than 2 mins to process 1 million data points (most of which is spend reading the input file!).

Value

A list containing input data (with detection score and feature indices) and a report (detected features).

Note

This algorihtm in now very robust - it is noise aware - and has several further modifications, improving its performance.

Author(s)

Tomas William Fitzgerald

References

Agilent

See Also

http://www.home.agilent.com/agilent/home.jspx

Examples

#exampleADM3()

Example output

ADM3 documentation built on May 29, 2017, 6:05 p.m.