Component: SPINLONG

SPINLONG

Identify user-defined spatio-temporal patterns in ChIP-seq and other sequencing data using the SPINLONG method. SPINLONG stands for Spatial Pattern Identification using Non-Linear OptimizatioN with Global constraints. Its applications include analysis of time-series RNA polymerase and histone modification data.

Version	1.0
Bundle	sequencing
Categories	Short-read Sequencing
Authors	Kristian Ovaska (kristian.ovaska@helsinki.fi)
Issue tracker	View/Report issues
Requires	commons-primitives-1.0.jar (jar) ; commons-math-2.1.jar (jar) ; csbl-javatools.jar (jar) ; jahmm-0.6.1.jar (jar) ; jcommon-1.0.16.jar (jar) ; jfreechart-1.0.13.jar (jar) ; saxon9he.jar (jar)
Source files	component.xml
Usage	Example with default values SPINLONG( bedFiles = BinaryFolder, patterns = XML, regions = CSV, //chromosomes = CSV, //mappability = BinaryFile, //control = BinaryFile, chromosomePattern = "", chromosomePreset = "hs37", fragmentSize = 200, maxDuplicateReads = 0, minRegionLength = 0, plotOptimizer = false, plotSegments = true, scoreThreshold = 0.1, seed = -1, threads = 4, yLog = false )

Version

1.0

Bundle

sequencing

Categories

Short-read Sequencing

Authors

Kristian Ovaska (kristian.ovaska@helsinki.fi)

Issue tracker

View/Report issues

Requires

commons-primitives-1.0.jar (jar) ; commons-math-2.1.jar (jar) ; csbl-javatools.jar (jar) ; jahmm-0.6.1.jar (jar) ; jcommon-1.0.16.jar (jar) ; jfreechart-1.0.13.jar (jar) ; saxon9he.jar (jar)

Source files

component.xml

Usage

Example with default values

Inputs

Name	Type	Mandatory	Description
bedFiles	BinaryFolder	Mandatory	Short reads in BED formats. The files may optionally by gzipped.
patterns	XML	Mandatory	XML file defining the patterns.
regions	CSV	Mandatory	Genomic regions.
chromosomes	CSV	Optional	Metadata for chromosomes. Must contain the columns Chromosome and Size, where Size gives the length of the chromosome. If not present, the parameter chromosomePreset is used.
mappability	BinaryFile	Optional	If given, contains genomic regions that are uniquely mappable in BED format. This is used to scale short read densities.
control	BinaryFile	Optional	If given, contains a control track in BED format that is subtracted from primary tracks (bedFiles input).

Name

Type

Mandatory

Description

bedFiles

BinaryFolder

Mandatory

Short reads in BED formats. The files may optionally by gzipped.

patterns

XML

Mandatory

XML file defining the patterns.

regions

CSV

Mandatory

Genomic regions.

chromosomes

CSV

Optional

Metadata for chromosomes. Must contain the columns Chromosome and Size, where Size gives the length of the chromosome. If not present, the parameter chromosomePreset is used.

mappability

BinaryFile

Optional

If given, contains genomic regions that are uniquely mappable in BED format. This is used to scale short read densities.

control

BinaryFile

Optional

If given, contains a control track in BED format that is subtracted from primary tracks (bedFiles input).

Outputs

Name	Type	Description
scores	CSV	Segment scores.
plots	ImageList	Segment plots and optionally optimizer plots. If both plotSegments and plotOptimizer are false, this output is empty.
patternsDump	HTML	Patterns formatted as HTML.

Name

Type

Description

scores

CSV

Segment scores.

plots

ImageList

Segment plots and optionally optimizer plots. If both plotSegments and plotOptimizer are false, this output is empty.

patternsDump

HTML

Patterns formatted as HTML.

Parameters

Name	Type	Default	Description
chromosomePattern	string	""	Java regular expression that selects the chromosomes to be used in analysis. The empty value selects all chromosomes. The pattern is matched against the first column of BED files.
chromosomePreset	string	"hs37"	If a custom chromosomes input is not given, this parameter is used to select a predefined chromosome set. Currently the legal values are hs36 (Homo sapiens, genome build 36) and hs37 (Homo sapiens, build 37).
fragmentSize	int	200	Size of sequences DNA fragments. The short reads are elongated so their final length matches this number. Setting this to 0 disables elongation.
maxDuplicateReads	int	0	The maximum number of duplicate short reads (same position and strand) that are utilized for each position. The rest are filtered out. This allows to renive reads that may be technical artefacts. If 0 or negative, all repeats are used.
minRegionLength	int	0	Minimum length for input regions that are processed. If 0, all regions are processed. This allows to filter out very short genes.
plotOptimizer	boolean	false	If true, visualize the progression of the optimizer. If false, omit optimizer plotting.
plotSegments	boolean	true	If true, visualize results of regions whose score is over the threshold. If false, omit plotting.
scoreThreshold	float	0.1	Minimum score for including a pattern match in the score and plot outputs. Notice that the score distribution depends on the scoring method.
seed	int	-1	Seed for random number generator. If negative, an automatically generated seed is used. Using a pre-defined seed ensures that results are deterministic.
threads	int	4	Maximum number of threads to use.
yLog	boolean	false	If true, the Y axis is plotted using logarithmic scale.

Name

Type

Default

Description

chromosomePattern

string

Java regular expression that selects the chromosomes to be used in analysis. The empty value selects all chromosomes. The pattern is matched against the first column of BED files.

chromosomePreset

string

"hs37"

If a custom chromosomes input is not given, this parameter is used to select a predefined chromosome set. Currently the legal values are hs36 (Homo sapiens, genome build 36) and hs37 (Homo sapiens, build 37).

fragmentSize

int

200

Size of sequences DNA fragments. The short reads are elongated so their final length matches this number. Setting this to 0 disables elongation.

maxDuplicateReads

int

The maximum number of duplicate short reads (same position and strand) that are utilized for each position. The rest are filtered out. This allows to renive reads that may be technical artefacts. If 0 or negative, all repeats are used.

minRegionLength

int

Minimum length for input regions that are processed. If 0, all regions are processed. This allows to filter out very short genes.

plotOptimizer

boolean

false

If true, visualize the progression of the optimizer. If false, omit optimizer plotting.

plotSegments

boolean

true

If true, visualize results of regions whose score is over the threshold. If false, omit plotting.

scoreThreshold

float

0.1

Minimum score for including a pattern match in the score and plot outputs. Notice that the score distribution depends on the scoring method.

seed

int

-1

Seed for random number generator. If negative, an automatically generated seed is used. Using a pre-defined seed ensures that results are deterministic.

threads

int

Maximum number of threads to use.

yLog

boolean

false

If true, the Y axis is plotted using logarithmic scale.

Test cases

Test case	Parameters▼	IN bedFiles	IN patterns	IN regions	IN chromosomes	IN mappability	IN control	OUT scores	OUT plots	OUT patternsDump
case1	properties	bedFiles	patterns	regions	(missing)	(missing)	(missing)	scores	(missing)	(missing)
	scoreThreshold=0, threads=1, maxDuplicateReads=2, seed=100
case2_chr	properties	bedFiles	patterns	regions	(missing)	(missing)	(missing)	scores	(missing)	(missing)
	scoreThreshold=0, threads=1, chromosomePattern=chr1, yLog=true
case3_xslt	properties	bedFiles	patterns	regions	(missing)	(missing)	(missing)	scores	(missing)	(missing)
	scoreThreshold=0.325, plotOptimizer=true
case4_xslt_simple	properties	bedFiles	patterns	regions	(missing)	(missing)	(missing)	scores	(missing)	(missing)
	scoreThreshold=0, plotSegments=false

Test case

Parameters▼

IN
bedFiles

IN
patterns

IN
regions

IN
chromosomes

IN
mappability

IN
control

OUT
scores

OUT
plots

OUT
patternsDump

case1