Overview of Help-Pages

ModelInspector: Search for Sequence Models

Introduction

ModelInspector uses a library of predefined models or models defined with FastM or FrameWorker to scan DNA sequences for matches to these models. A model consists of various individual elements (like transcription factor binding sites, repeats, hairpins), their strand orientation, their sequential order, and their distance ranges.

ModelInspector uses a proprietary scoring algorithm to allow inclusion of very different element types into the composite scoring of matches. Thus, IUPAC sequence elements can be successfully combined with different types of weight matrices and structural elements (e.g. hairpins) in the assessment of match quality.

The ModelInspector and FastM algorithm is described in Frech et al., 1997 (JMB), and Klingenhoff et al., 1999 (Bioinformatics).

The following predefined libraries are available:

• the Promoter Module Library
  This library includes models for functional subunits of promoters. All promoter modules are experimentally verified.
• the Genomic Repeat Library
  This library includes models for ALUs, L1 elements, THEs, and B1 elements.
• the Long Terminal Repeat Library
  This library includes models for Lentivirus LTRs and avian and mammalian C-type LTRs.

All models of the Genomic Repeat and Long Terminal Repeat Library show a very high specificity.

Input

Database input

Select one of these database-sections

On the Genomatix server the following databases are available: Genomatix Promoter Database: Promoters of annotated genes Subset of all human, mouse, and rat promoters. Promoters of hypothetical proteins (e.g Loc127262) or genes that are annotated as "similar to ..." (e.g. Loc419384) are omitted. Genomatix Promoter Database: Promoters of all genes All promoter sequences extracted fromElDoradogenomes with "Genomatix optimized length" (500 bp upstream of the first TSS and 100 bp downstream of the last TSS). Genomatix ElDorado Genomes All genomes available inElDorado(human, mouse, rat, chimpanzee, rhesus monkey, dog, opossum, platypus, cow, horse, chicken, zebrafish, fruitfly, Anopheles, honeybee, C. elegans, Arabidopsis and rice) Other databases Philipp Bucher's Eukaryotic Promoter Database (EPD) NCBI Reference Sequences (mRNA sequences) GenBank sections The sections Bacteria, Invertebrates, Other Mammalian, Other Vertebrates, Plants, Primates, Rodents, and Viral are available. In case you have selected a section from the GenBank database you may also restrict the analysis to sequences containing user-defined keywords in their annotation. You can enter keywords which will be searched in the keyword line of the annotation the description line of the annotation the complete annotation The keyword searches can be combined with "AND" or "OR". Please note that the keywords cannot contain blanks (all blanks will be skipped). These parameters are hidden by default. You can use the next to the section header to reveal them!

ModelInspector Parameters

Model Library Selection
Library version	Here you can select a previous version of the promoter module library. This can be helpful for re-producing old results. By default, the latest promoter module library is selected (please see the Module Library Release Notes). Each version of the module library corresponds to a specific version of the matrix library (the matrix library version with which the module definitions have been generated): Module Library Version Matrix Library Version Module Library 5.4 Matrix Library 8.3 Module Library 5.3 Matrix Library 8.2 Module Library 5.2 Matrix Library 8.1 Module Library 5.1 Matrix Library 8.0 Module Library 5.0 Matrix Library 7.1 Module Library 4.5 Matrix Library 7.0 Module Library 4.4 Matrix Library 6.3 Module Library 4.3 Matrix Library 6.2 Module Library 4.2 Matrix Library 6.1 The module library version selected also affects searches for user-defined models (i.e. the matrix library version corresponding to the module library version selected is used for the model searches). It is necessary to change the module library version (and thus the matrix library version) when user-defined models contain matrix families that have been removed or renamed in newer library versions.
Model groups	Please choose one or several of the available Genomatix model libraries. If you have created your own models with FastM or FrameWorker, they can be found in the "User-defined models" library. You can decide if you want to use all the models in the chosen libraries use previously defined model subsets or continue with a subset selection. In the third case, there will be a separate page with a list of all models in the chosen libraries and you can select your model subset by clicking the checkboxes for each model. If you started ModelInspector directly from a previous FastM session, the list will only contain the model you just created with FastM.
Search Parameters
Max. number of matches	Enter the maximum number of matches in the output file. In case the output is filtered for matches occurring in selected annotated sequence regions, only the filtered matches are considered for the maximum number of matches. Hint for user-defined models: If you find a lot of matches in the first few sequences of a database section (e.g. most sequence names with matches starting with A or B), you might want to change your model to be more specific (e.g. raise the matrix similarities for binding sites, thus defining a higher specificity for the binding site search; or restrict the strand orientation to the sense or antisense strand). In addition, you can supply more specific (i.e. smaller) distance ranges.
Threshold	Enter a threshold for the output of model matches. This value gives the minimum score that a match has to reach to appear in the output file. The value is given in percent of the number of individual elements of the model. Default is 100 % (i.e. all elements of the model have to be present). These parameters are hidden by default. You can use the next to the section header to reveal them!
Strand	If this option is checked, only the top strand of the input or database sequences is scanned for model matches. Per default both strands are searched. These parameters are hidden by default. You can use the next to the section header to reveal them!
Annotation filter	Generally, all matches are listed in the output. Alternatively, the output can be filtered for matches located in regions annotated as 3'UTR regions annotated as 5'UTR regions annotated as exons regions annotated as introns regions annotated as promoters regions annotated as repeats unannotated sequence regions Note: The annotation filter is only available for databases where the above mentioned features are annotated in the feature table of the database entries, i.e. the filter can be used for ElDorado genomes, EMBL and GenBank database sections. These parameters are hidden by default. You can use the next to the section header to reveal them!
Ranking	In case one of the Genomatix promoter databases is scanned with a model, the search results are evaluated by calculation of p-values for Gene Ontology groups. This evaluation shows whether genes identified by the model are functionally related. As the Gene Ontology ranking takes some time, it can be switched off.
Output Parameters
Offset for match positions	Enter an offset (in number of basepairs, can also be negative) that will be added to each position in the output file. This feature can be used i.e. in cases where the transcription start site is known and positions should be given relative to the TSS. E.g. if the TSS is at position 500 in a sequence, the offset should be "-500" for relative positions. These parameters are hidden by default. You can use the next to the section header to reveal them!
Statistics	If the statistics option is checked, only the statistics output is created. In this case the number of matches is unlimited. This option is useful if you are interested in the total number of matches in a large data set (e.g. the Human Genome) as the number of matches shown in the match overview is limited. These parameters are hidden by default. You can use the next to the section header to reveal them!
Alternative matches	If the alternative matches option is checked, alternative model matches are displayed additionally in the detailed output. For overlapping model matches (i.e. model matches where the start and end positions are identical or differ by less than 5 base pairs), only the model match with the highest score of individual elements is shown. The alternative model matches can be displayed optionally in the detailed output in order to check the positions and scores of the individual elements. These parameters are hidden by default. You can use the next to the section header to reveal them!
Output sorted by	The output can be sorted by the name of your models by the quality of the model matches (best first) or by sequence position (first match on the sequence first) These parameters are hidden by default. You can use the next to the section header to reveal them!
Output filtered for	The output can be filtered for sequences in which at least the specified number of different model matches occur. Per default, all sequences with at least one model match are shown in the output. These parameters are hidden by default. You can use the next to the section header to reveal them!
Email address	Here you can choose between two methods for receiving the results: Show result directly in browser window In this option the URL of the result is directly shown in your browser window. Warning: Please use this option only for analyses which can be performed in a short time. If the analysis takes longer than the timeout of the webserver, the connection will be terminated and you will receive an error message (e.g. "The document contained no data."). In this case, the results will not be available, please restart the analysis using the option below "Send the URL of the result to". Send the URL of the result via email In this option an email with the URL of the results will be sent to the user provided email address, when the analysis is finished. The results will be available for a limited time on our server. For details of how long your results will be kept please see the result-email. After that period they will be deleted unless protected in the project management!

ModelInspector Output

ModelInspector generates three output files, the match overview, the detailed output, and the statistics file.

1. Match overview

The first output file of ModelInspector contains:

• The selected input parameters.
• A match list with the position and strand orientation of each model match.
  If one of the Genomatix promoter databases or ElDorado genomes is searched, ElDorado links to the genomic location of the matches are provided.
• In case one of the Genomatix promoter databases is scanned with a model, the search results are evaluated by calculation of p-values for Gene Ontology groups (categories "biological process", "molecular function", and "cellular component"). This evaluation shows whether the genes identified by the model are functionally related.
• The model matches can be extracted in several ways, e.g. you can extract the sequence of the matches. Please select the matches in the match table by checking the select boxes in the column " Select Match" or by using the buttons below the match table. Per default, all matches are selected.

  Extraction Options
  Sequence Extraction	You can extract the complete sequence or only the match positions (default) It is possible to extract further sequence positions upstream and downstream of the match positions. In this case you have to enter the number of additional basepairs into the box. Default value is ± 0 basepairs.
  GeneID Extraction	The button "Extract GeneIDs" extracts the GeneIDs of the matching sequences for each model separately. The extracted GeneIDs can be used e.g. as input for GePS. The button "Extract GeneIDs by Chromosome" extracts the GeneIDs of the matching sequences for each chromosome separately.
  Match Extraction	The button "Export Matches in EXCEL format" allows to export all information available in the match overview (like model name, sequence name and position of the model match) to a tab-delimited file. This file is saved to your local disk and can be opened directly with Microsoft Excel. The button "Export Matches in BED file format" allows to export genomic model matches to a BED file (e.g. for upload into RegionMiner). The button is only available when a genome search has been performed or when the input sequences include chromosomal positions in their annotation.

• In case you have searched one of the Genomatix promoter databases (where the genes are identified by their GeneIDs) you can compare the result with a list of GeneIDs (e.g. from another ModelInspector search).

  Compare results
  Enter GeneIDs	Enter a list of GeneIDs separated by spaces, returns, or commas.

  When you press the "Compare" button you will get the information which GeneIDs are common and which GeneIDs are specific either for your input list or for the ModelInspector result.

• In case of database searches the model matches can be further evaluated by a PubMed search.

  Further Evaluation of Matches
  Search PubMed for	PubMed can be searched for entries containing the gene name which is automatically extracted from the description line of the matching sequences and further keywords that can be entered by the user. Default for these keywords is ("promoter" OR "transcription factor").

  When you press the "Extract gene names" button you will receive a list with the extracted gene names and links to the corresponding PubMed queries.

  Note: The gene name extraction works only for genomic eukaryotic DNA sequences.

Example for the match overview:

Example for evaluation of results:

For this example, the experimentally verified module "CDEF_CHRF_01" from the Genomatix Promoter Module Library which is involved in cell cycle regulation was searched in all human promoters.

2. Detailed output

The second output file of ModelInspector contains detailed information for each individual element of the model:

• positions on the sequence and strand information
• features of the element like sequence, length, matrix similarity, loop length, etc.

Example for the detailed match list:

Inspecting sequence humactga [M19283] (1 - 575):

Model: YY1F_SRFF_01 (378 - 397 (+))

3. Statistics

The third output file of ModelInspector contains a statistics of the model matches and detailed information for your own models:

• Statistics of the number of matches for each model
• Detailed statistics for your own models, i.e. statistics of the found number of matches for each individual element and minimum and maximum distance to the following element
  '(+) strand matches' denotes the number of matches taken into account for model matches on the sense strand, '(-) strand matches' the number of matches considered for model matches on the antisense strand.

Example for the statistics:

References

If you are interested in more details, ModelInspector and FastM are described in

• Frech K, Danescu-Mayer J, Werner T (1997).
  A novel method to develop highly specific models for regulatory units detects a new LTR in GenBank which contains a functional promoter.
  J. Mol. Biol. 270 (5), 674-687.
  
  
• Klingenhoff A, Frech K, Quandt K, Werner T (1999).
  Functional promoter modules can be detected by formal models independent of overall nucleotide sequence similarity.
  Bioinformatics 15, 180-186.