affyread - Read microarray data from Affymetrix GeneChip file

Syntax

AffyStruct = affyread(File) AffyStruct = affyread(File, LibraryPath)

Description

Note This function does not work on the Solaris platform.

AffyStruct = affyread(File) reads an Affymetrix file and creates a MATLAB structure. The affyread function can read Affymetrix EXP, DAT, CEL, CLF, BGP, CDF, and GIN files associated with Affymetrix GeneChip arrays for expression, genotyping (SNP), or resequencing assays. It can read Affymetrix CHP files associated with Affymetrix GeneChip arrays for expression assays only.

AffyStruct = affyread(File, LibraryPath) specifies the path and directory of a CDF or GIN library file.

Reading many CEL files and/or a large CEL file can require extended amounts of memory from the operating system. If you receive any errors related to memory or have trouble reading CEL files, try the following:

Increase the virtual memory (swap space) for your operating system (with a recommended initial size of 3,069 and a maximum size of 16,368) as described at:
```
 http://www.mathworks.com/support/tech-notes/1100/1106.html#6
```
Set the 3-GB switch (32-bit Windows^® XP only) as described at:
```
 http://www.mathworks.com/support/tech-notes/1100/1107.html
```

For help resolving errors related to memory, see Resolving "Out of Memory" Errors.

Inputs

File

String specifying a file name or a path and file name of one of the following Affymetrix file types associated with Affymetrix GeneChip arrays for expression, genotyping (SNP), or resequencing assays. However, if the file name is for a CHP file, it must be associated with an Affymetrix GeneChip array for an expression assay.

EXP — Data file containing information about experimental conditions and protocols.
DAT — Data file containing raw image data (pixel intensity values).
CEL — Data file containing information about the intensity values of the individual probes.
CHP — Data file containing summary information of the probe sets, including intensity values.
CLF — Cell layout file that maps probe IDs to a location (x- and y-coordinates) in the CEL file.
BGP — Background probe file that lists the probes to use for background correction.
CDF — Library file containing information about which probes belong to which probe set.
GIN — Library file containing information about the probe sets, such as the gene name associated with the probe set.

If you specify only a file name, put that file on the MATLAB search path or in the current directory. If you specify only a file name of a CDF or GIN library file, you can specify the path and directory in the LibraryPath input argument.

Tip You can learn more about the Affymetrix GeneChip files and download sample files from:

http://www.affymetrix.com/support/technical/sample_data/demo_data.affx

Note Some Affymetrix sample data files (DAT, EXP, CEL, and CHP) are combined in a DTT or CAB file. Download and use the Affymetrix Data Transfer Tool to extract these files from the DTT or CAB file. You can download the Data Transfer Tool from:

http://www.affymetrix.com/products/software/specific/dtt.affx

You will have to register and log in at the Affymetrix Web site to download the Data Transfer Tool.

LibraryPath

String specifying the path and directory of a:

CDF library file associated with File when File is a CHP file
CDF library file when File is a CDF file
GIN library file when File is a GIN file

Note If you do not specify LibraryPath when reading a CHP file, affyread looks in the current directory for the CDF file. If it does not find the CDF file, it still reads the CHP file. However, it omits the probe set names and types from the return value, AffyStruct.

Outputs

AffyStruct

MATLAB structure containing information from an Affymetrix data or library file, for expression, genotyping (SNP), or resequencing assay types.

The following tables describe the fields in AffyStruct for the different Affymetrix file types.

EXP, DAT, CEL, CHP, CLF, BGP, CDF, and GIN Files

Field	Description
`Name`	File name.
`DataPath`	Path and directory of the file.
`LibPath`	Path and directory of the CDF and GIN library files associated with the file you are reading.
`FullPathName`	Path and directory of the file.
`ChipType`	Name of the AffymetrixGeneChip array (for example, DrosGenome1 or HG-Focus).
`Date` or `CreateDate`	File creation date.

EXP File

Field	Description
`ChipLot` `Operator` `SampleType` `SampleDesc` `Project` `Comments` `Reagents` `ReagentLot` `Protocol` `Station` `Module` `HybridizeDate` `ScanPixelSize` `ScanFilter` `ScanDate` `ScannerID` `NumberOfScans` `ScannerType` `NumProtocolSteps` `ProtocolSteps`	Information about experimental conditions and protocols captured by the Affymetrix software.

DAT File

Field	Description
`NumPixelsPerRow`	Number of pixels per row in the image created from the GeneChip array (number of columns).
`NumRows`	Number of rows in the image created from the GeneChip array.
`MinData`	Minimum intensity value in the image created from the GeneChip array.
`MaxData`	Maximum intensity value in the image created from the GeneChip array.
`PixelSize`	Size of one pixel in the image created from the GeneChip array.
`CellMargin`	Size of gaps between cells in the image created from the GeneChip array.
`ScanSpeed`	Speed of the scanner used to create the image.
`ScanDate`	Date the scan was performed.
`ScannerID`	Name of the scanning device used.
`UpperLeftX` `UpperLeftY` `UpperRightX` `UpperRightY` `LowerLeftX` `LowerLeftY` `LowerRightX` `LowerRightY`	Pixel coordinates of the scanned image.
`ServerName`	Not used.
`Image`	A `NumRows`-by-`NumPixelsPerRow` image of the scanned GeneChip array.

CEL File

Field	Description
`FileVersion`	Version of the CEL file format.
`Algorithm`	Algorithm used in the image-processing step that converts from DAT format to CEL format.
`AlgParams`	String containing parameters used by the algorithm in the image-processing step.
`NumAlgParams`	Number of parameters in `AlgParams`.
`CellMargin`	Size of gaps between cells in the image created from the GeneChip array, used for computing the intensity values of the cells.
`Rows`	Number of rows of probes.
`Cols`	Number of columns of probes.
`NumMasked`	Number of masked probes, which are not used in subsequent processing.
`NumOutliers`	Number of cells identified as outliers (extremely high or extremely low intensity) by the image-processing step.
`NumProbes`	Number of probes (`Rows` * `Cols`) on the GeneChip array.
`UpperLeftX` `UpperLeftY` `UpperRightX` `UpperRightY` `LowerLeftX` `LowerLeftY` `LowerRightX` `LowerRightY`	Pixel coordinates of the scanned image.
`ProbeColumnNames`	Cell array containing the eight column names in the `Probes` field: `PosX` — x-coordinate of the cell `PosY` — y-coordinate of the cell `Intensity` — Intensity value of the cell `StdDev` — Standard deviation of intensity value `Pixels` — Number of pixels in the cell `Outlier` — True/false flag indicating if the cell was marked as an outlier `Masked` — True/false flag indicating if the cell was masked `ProbeType` — Integer indicating the probe type (for example, `1` = expression)
`Probes`	`NumProbes`-by-8 array of information about the individual probes, including intensity values. The `ProbeColumnNames` field contains the column names of this array.

CHP File

Field	Description
`AssayType`	Type of assay associated with the GeneChip array (for example, Expression, Genotyping, or Resequencing).
`CellFile`	File name of the CEL file from which the CHP file was created.
`Algorithm`	Algorithm used to convert from CEL format to CHP format.
`AlgVersion`	Version of the algorithm used to create the CHP file.
`NumAlgParams`	Number of parameters in `AlgParams`.
`AlgParams`	String containing parameters used in steps required to create the CHP file (for example, background correction).
`NumChipSummary`	Number of entries in `ChipSummary`.
`ChipSummary`	Summary information for the GeneChip array, including background average, standard deviation, max, and min.
`BackgroundZones`	Structure containing information about the zones used in the background adjustment step.
`Rows`	Number of rows of probes.
`Cols`	Number of columns of probes.
`NumProbeSets`	Number of probe sets on the GeneChip array.
`NumQCProbeSets`	Number of QC probe sets on the GeneChip array.
`ProbeSets` (Expression GeneChip array)	`NumProbeSets`-by-1 structure array containing information for each expression probe set, including the following fields: `Name` — Name of the probe set. `ProbeSetType` — Type of the probe set. `CompDataExists` — True/false flag indicating if the probe set has additional computed information. `NumPairs` — Number of probe pairs in the probe set. `NumPairsUsed` — Number of probe pairs in the probe set used for calculating the probe set signal (not masked). `Signal` — Summary intensity value for the probe set. `Detection` — Indicator of statistically significant difference between the intensity value of the PM probes and the intensity value of the MM probes in a single probe set (`Present`, `Absent`, or `Marginal`). `DetectionPValue` — P-value for the `Detection` indicator. `CommonPairs` — When `CompDataExists` is `true`, contains the number of common pairs between the experiment and the baseline after the removal of outliers and masked probes. `SignalLogRatio` — When `CompDataExists` is `true`, contains the change in signal between the experiment and baseline. `SignalLogRatioLow` — When `CompDataExists` is `true`, contains the lowest ratios of probes between the experiment and the baseline. `SignalLogRatioHigh` — When `CompDataExists` is `true`, contains the highest ratios of probes between the experiment and the baseline. `Change` — When `CompDataExists` is `true`, describes how the probe changes versus a baseline experiment. Choices are `Increase`, `Marginal Increase`, `No Change`, `Decrease`, or`Marginal Decrease`. `ChangePValue` — When `CompDataExists` is `true`, contains the p-value associated with `Change`.
`ProbeSets` (Genotyping GeneChip array)	`NumProbeSets`-by-1 structure array containing information for each genotyping probe set, including the following fields: `Name` — Name of the probe set. `AlleleCall` — Allele that is present for the probe set. Possibilities are `AA` (homozygous for the major allele), `AB` (heterozygous for the major and minor allele), `BB` (homozygous for the minor allele), or `NoCall` (unable to determine allele). `Confidence` — Measure of the accuracy of the allele call. `RAS1` — Relative Allele Signal 1 for the SNP site, which is calculated using sense probes. `RAS2`— Relative Allele Signal 2 for the SNP site, which is calculated using antisense probes. `PValueAA` — p-value for an `AA` call. `PValueAB` — p-value for an `AB` call. `PValueBB` — p-value for a `BB` call. `PValueNoCall` — p-value for a `NoCall` call.
`ProbeSets` (Resequencing GeneChip array)	`NumProbeSets`-by-1 structure array containing information for each resequencing probe set, including the following fields: `CalledBases` — 1-by-`NumProbeSets` character array containing the bases called by the resequencing algorithm. Possible values are `a`, `c`, `g`, `t`, and `n`. `Scores` — 1-by-`NumProbeSets` array containing the score associated with each base call.

CLF File

Field Description

LibSetName Name of a collection of related library files for a given chip. There is only one LibSetName for a CLF file. For example, PGF and CLF files intended for use together must have the same LibSetName.

LibSetVersion Version of a collection of related library files for a given chip. There is only one LibSetVersion for a CLF file. For example, PGF and CLF files intended for use together must have the same LibSetVersion.

GUID Unique identifier for the CLF file.

CLFFormatVersion Version of the CLF file format.

Rows

Number of rows in the CEL file.

Note The CLF file is 1 base, which means the first row and column are designated 1,1, not 0,0.

Cols

Number of columns in the CEL file.

Note The CLF file is 1 base, which means the first row and column are designated 1,1, not 0,0.

StartID

Starting number for the numbering of elements in the CLF file.

Tip This information is useful when numbering does not start with 1.

EndID

Ending number for the numbering of elements in the CLF file.

Tip This information is useful when numbering does not start with 1 and/or there are gaps in the numbering.

Order Order in which the probe IDs are numbered in the CEL file, either 'row_major' or 'col_major'.

DataColNames Names of the columns in the CEL file that contain data.

Data

If the numbering of elements in the CLF file is sequential, this field contains a function handle that calculates the x- and y- coordinates of each element in the file from the probe ID.

If the numbering of elements in the CLF file is not sequential, this field contains a matrix indicating the number value of each element in the file.

BGP File

Field	Description
`LibSetName`	Name of a collection of related library files for a given chip. There is only one `LibSetName` for a BGP file.
`LibSetVersion`	Version of a collection of related library files for a given chip. There is only one `LibSetVersion` for a BGP file.
`GUID`	Unique identifier for a BGP file.
`ExecGUID`	Information about the algorithm used to generate the BGP file.
`ExecVersion`
`Cmd`
`Data`	Structure containing the following fields: `probe_id` — ID of the probe to use for background correction. `probeset_id` — ID of the probe set in the PGF file to which the probe belongs. `type` — Classification information for the probe. `gc_count` — Combined number of G and C bases in the probe. `probe_length`— Length of the probe in base pairs. `interrogation_position` — Interrogation position of the probe. It is typically 13 for 25-mer PM/MM probes. `probe_sequence` — Sequence of the probe on the array, going in the direction from array surface to solution. For most standard Affymetrix arrays, this direction is from 3' to 5'. For example, for a sense target (st) probe (see the `probe_type` field), complement the sequence in this field before looking for matches to transcript sequences. For an antisense target (at), reverse this sequence. `atom_id` — ID of the atom to which the probe belongs. `x` — Column coordinate of the probe in the CEL file. `y` — Row coordinate of the probe in the CEL file. `probeset_type` — Classification information for the probe set, such as control, affx, or spike. This type information can include multiple classifications and can also be nested. `probe_type` — Classification information for the probe, such as pm (perfect match), mm (mismatch), st (sense target), or at (antisense target). This type information can include multiple classifications and can also be nested.

CDF File

Field	Description
`Rows`	Number of rows of probes.
`Cols`	Number of columns of probes.
`NumProbeSets`	Number of probe sets on the GeneChip array.
`NumQCProbeSets`	Number of QC probe sets on the GeneChip array.
`ProbeSetColumnNames`	Cell array containing the six column names in the `ProbePairs` field in the `ProbeSets` array: `GroupNumber` — Number identifying the group to which the probe pair belongs. For expression arrays, this value is always `1`. For genotyping arrays, this value is typically `1` (allele A, sense), `2` (allele B, sense), `3` (allele A, antisense), or `4` (allele B, antisense). `Direction` — Number identifying the direction of the probe pair. `1` = sense and `2` = antisense. `PMPosX` — x-coordinate of the perfect match probe. `PMPosY` — y-coordinate of the perfect match probe. `MMPosX` — x-coordinate of the mismatch probe. `MMPosY` — y-coordinate of the mismatch probe.
`ProbeSets`	`NumProbeSets`-by-1 structure array containing information for each probe set, including the following fields: `Name` — Name of the probe set. `ProbeSetType` — Type of the probe set. `CompDataExists` — True/false flag indicating if the probe set has additional computed information. `NumPairs` — Number of probe pairs in the probe set. `NumQCProbes` — Number of QC probes in the probe set. `QCType` — Type of QC probes. `GroupNames` — Name of the group to which the probe set belongs. For expression arrays, this field contains the name of the probe set. For genotyping arrays, this field contains the name of the alleles, for example `{'A' 'C' 'A' 'C'}'`. `ProbePairs` — `NumPairs`-by-6 array of information about the probe pairs. The column names of this array are contained in the `ProbeSetColumnNames` field.

GIN File

Field	Description
`Version`	GIN file format version.
`ProbeSetName`	Probe set ID/name.
`ID`	Identifier for the probe set (gene ID).
`Description`	Description of the probe set.
`SourceNames`	Source or sources of the probe sets.
`SourceURL`	Source URL or URLs for the probe sets.
`SourceID`	Vector of numbers specifying which `SourceNames` or `SourceURL` each probe set is associated with.

Examples

The following example uses the demo data and CDF library file from the E. coli Antisense Genome array, which you can download from:

http://www.affymetrix.com/support/technical/sample_data/demo_data.affx

After downloading the demo data, you need the Affymetrix Data Transfer Tool to extract the CEL, DAT, and CHP files from a DTT file. You can download the Data Transfer Tool from:

http://www.affymetrix.com/products/software/specific/dtt.affx

The following example assumes that you have stored the files Ecoli-antisense-121502.CEL, Ecoli-antisense-121502.dat, and Ecoli-antisense-121502.chp on the MATLAB search path or in the current directory. It also assumes that you have stored the associated CDF library file, Ecoli_ASv2.CDF, at D:\Affymetrix\LibFiles\Ecoli.

Read the contents of a CEL file into a MATLAB structure.
```
celStruct = affyread('Ecoli-antisense-121502.CEL');
```
Display a spatial plot of the probe intensities.
```
maimage(celStruct, 'Intensity')
```
Zoom in on a specific region of the plot.
```
axis([200 340 0 70])
```
Read the contents of a DAT file into a MATLAB structure. Display the raw image data, and then use the axis image function to set the correct aspect ratio.
```
datStruct = affyread('Ecoli-antisense-121502.dat');
imagesc(datStruct.Image)
axis image
```
Zoom in on a specific region of the plot.
```
axis([1900 2800 160 650])
```

Read the contents of a CHP file into a MATLAB structure, specifying the location of the associated CDF library file. Then extract information for probe set 3315278.

chpStruct = affyread('Ecoli-antisense-121502.chp',...
            'D:\Affymetrix\LibFiles\Ecoli');
geneName = probesetlookup(chpStruct,'3315278')

geneName = 

      Identifier: '3315278'
    ProbeSetName: 'argG_b3172_at'
        CDFIndex: 5213
        GINIndex: 3074
     Description: [1x82 char]
          Source: 'NCBI EColi Genome'
       SourceURL: [1x74 char]