quantify: Quantification of results

In robinweide/GENOVA: GENOVA: expore the Hi-C

Quantification of results

Description

A good amount of discovery objects can be quantified. What exactly is to be quantified differs per discovery type.

saddle disoveries

can be used to compute compartment strengths.

ARMLA discoveries

such as APA, PE-SCAn, ATA and ARA compare different regions of their outputs.

IIT discoveries

summarise their values by neighbours.

Usage

quantify(discovery, ...)

## S3 method for class 'APA_discovery'
quantify(
  discovery,
  size = 3,
  metric = "median",
  shape = "center_vs_quadrants",
  IDX = NULL,
  ...
)

## S3 method for class 'PESCAn_discovery'
quantify(
  discovery,
  size = 5,
  metric = "median",
  shape = "circle",
  IDX = NULL,
  ...
)

## S3 method for class 'CSCAn_discovery'
quantify(
  discovery,
  size = 5,
  metric = "median",
  shape = "circle",
  IDX = NULL,
  ...
)

## S3 method for class 'ATA_discovery'
quantify(
  discovery,
  size = 3,
  metric = "median",
  shape = "insulation",
  IDX = NULL,
  ...
)

## S3 method for class 'ARA_discovery'
quantify(discovery, size = 3, shape = "ARA", ...)

## S3 method for class 'saddle_discovery'
quantify(discovery, ...)

## S3 method for class 'IIT_discovery'
quantify(discovery, ...)

Arguments

discovery	A discovery object as returned by GENOVA analysis functions.
...	further arguments passed to or from other methods. take the middle 3x3 matrix of the APA).
size	An integer of length one to determine the size of features of interest in bins.
metric	Either "median" or "mean" to summarise features.
shape	A character of length 1 specifying what shape to use. See the section shapes.
IDX	The IDX part of a contacts object. Used only in converting features expressed in bins back to genomic space. This is rarely needed, but is useful for APAs ran with extended loops where features aren't 1:1 traceable to the input.

Shapes

The quantification of ARMLA discoveries require a shape to distinguish regions to quantify.

ARA, PESCAn and CSCAn

APA, PESCAn and CSCAn require one of the following:

• "center_vs_quadrants"

• "center_vs_rest"

• "circle"

The size parameter determines the number of bins of the central foreground.

In the illustrations below, red is considered foreground and blue is considered background.

The "center_vs_quadrants" option does not include region directly horizontal or vertical of the centre as background.

The "center_vs_rest" option sees everything but the centre as background.

The "circle" option is like "center_vs_rest" but rounds corners of the central foreground. Note that for size <= 3 these two options are equivalent.

ATA

ATA requires one of the following:

• "insulation"

• "cornerpeak"

• "checker"

In the illustrations below, red is considered foreground and blue is considered background. The line indicates the diagonal.

The "insulation" option compares within-TAD contacts to between-TAD contacts.

The "cornerpeak" option compares the intersection of boundaries versus within-TAD contacts.

The "checker" option compares the within-TAD contacts to the contacts between it's immediate neighbours.

ARA

ARA requires one of the following:

• "ARA"

• "stripes"

The size argument controls the width of the stripes.

The "ARA" option reports about both 3' and 5' stripes and regions as well as the bins that span the locus. They are indicated in different colours below.

The "stripes" options reports the values and distances of the stripes. The 5' distances are encoded as negative, whereas 3' distances are positive.

Examples

NULL

robinweide/GENOVA documentation built on March 14, 2024, 11:16 p.m.