Data files contain a user's own experimental data - the list of genes that were up- or down-regulated in a microarray experiment. Genes that were up-regulated and genes that were down-regulated should be put in separate files and analyzed separately. The file is simply a list of unique probe identifiers for the particular microarray system that was used, one identifier per line:

probeID1
probeID2
probeID3

Support for a variety of popular microarray systems is built-in to L2L. If your microarray system isn't among them, you can create your own translator library. The following table lists the supported microarray systems and a few sample probe identifiers. Note that probe identifiers for chip sets (U133 Set, U95 Set, etc.) include the chip ID. All identifiers are case-insensitive (i.e. 200007_at or 200007_AT are both fine).

Built-in microarray platforms and sample probe identifiers

"All HUGO Names" is intended to be used for gene annotation - you can put a few genes you want to annotate in your data file, use this translator library, and see which L2L Microarray Database lists your genes of interest are found on. "All HUGO Names" can also be a used as a default "microarray system" if your microarray is not represented and you do not want to create a translator library for it. However, L2L's statistical analysis relies on knowing how many genes are actually on your microarray, and how many of those were changed in your experiment. Therefore, you should not put much faith in any P values or fold-enrichment numbers if you use "All HUGO Names". It is also very easy to create a new translator library (see below), so we highly recommend you do this if L2L doesn't include a translator library for your microarray system.

A translator library allows L2L to translate gene names to microarray probe identifiers and back. It is a tab-delimted file with a paired probe identifer and HUGO name on each line:

#L2L     library
#NAME    Affy-SomeArray_v1
#RELEASE  2007.1
probeID1    XYZ1
probeID2    ABCD1
probeID3    HUJA6

The annotations at the top of the file are optional; they are not used by L2L. The probe identifers can be anything, as long as they match the probe identifiers you use for your data. Try to avoid special characters, however. The web-interface of L2L will warn you if your uploaded translator library has improper characters in it (this is a security measure). Gene names must be official HUGO gene symbols in order for L2L's gene annotation functions to work (linking to EntrezGene, for example).

Each list in the database is a file with a few annotations at the top, followed by the HUGO gene symbols of all the genes on that list, one per line:

#L2L     listfile
#NAME    brca1_overexp_up
#REFERENCE    12032322
#DESCRIPTION  Upregulated by induction of BRCA1 in EcR-293 cells
#KEYWORDS cancer
#PLATFORM HuGeneFL
#RELEASE  2007.1
#ORGANISM human
#REPOSITORY 
FSTL1
GALNT3
SEC10L1
HTATIP...

The first line tells the L2L application that this is, indeed, a list. This line should be the same for all list files. The second line is a short, informative name for the list (the same as the file name). It should contain only alphanumeric characters and underscores. The third line is a reference to the source of the list. For L2L Microarray Database lists, this is the PubMed ID of the source publication. The fourth line is a description of the list. It can be as long as necessary, and can include any character except tabs.

The fifth line can contain one of a number of keywords for browsing the database and (in a future revision of L2L) restricting searches to particular topics. Current keywords. The sixth line describes the platform (microarray or otherwise) that was used to generate the data encompassed by the list. The seventh line is the release version of the list; this is for the user's reference only, and is not used by the L2L program (i.e. it can be omitted). The eight line is the organism on from which the data was generated. The ninth line contains a repository accession number, if this publication was linked to any data in Gene Expression Omnibus or Array Express. All other lines in the file contain one of the genes on the list. Note that the order and position of the lines is not critical; L2L identifies annotations by the "#XXX" designators.

Beginning with release 2006.2 (L2L v1.1), the L2L command-line tool can use a single-file database format to speed high-throughput batch processing. This format is simply a concatenation of all list files into a single text file, with a line containing "##" at the end of each list of genes.

L2L provides a variety of annotations and hyperlinks to external references in its output HTML files. The data needed to generate these links are contained in a text file, "listsets.txt", with a format similar to the single-file databases:

#L2L     listset
#NAME    l2lmdb
#REFERENCE  http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=pubmed&dopt=Abstract&list_uids=XURLX
#REFDESC    PubMed Abstract
#DESCRIPTION  L2L MDB
##

"NAME" is the shorthand name to be used in file and folder names. "REFERENCE" is a template for hyperlinking to the data source for a list in this database; "XURLX" will be replaced with the "REFERENCE" annotation from the individual list. "REFDESC" is a description of the hyperlink destination for display in output web pages. "DESCRIPTION" is also for display in output web pages (a longhand version of "NAME"). "##" marks the end of one entry in the file.