The results of the analyses run by Genotator can be viewed with the interactive graphical browser. The browser displays color-coded sequence annotations on a canvas that can be scrolled and zoomed, allowing the annotated sequence to be explored at multiple levels of detail. The user can view the actual DNA sequence in a separate window; when a region is selected in the map display, it is automatically highlighted in the sequence display, and vice-versa. By displaying the output of all of the sequence analyses, Genotator provides an intuitive way to identify the significant regions (for example, probable exons) in a sequence. Users can interactively add personal annotations to label regions of interest. Additional capabilities of Genotator include primer design and pattern searching. A new feature in version 1.51 of the browser lets you retrieve the GenBank record for a sequence that has significant homologies to your sequence.
Genotator runs on UNIX workstations. The back end is written in perl and Tkperl, and calls out to various sequence analysis programs written here and elsewhere. The front end, which is also written in perl and Tkperl, uses Gregg Helt's bioTkperl widgets.
The Genotator back end can be invoked via the graphical user interface (GUI) or with command-line options. To invoke Genotator via the GUI, type
~nomi/genotator/genotator
(for local LBNL users; offsite users will use a different directory).
Click the file selection box (here labeled "humtfpb" because the user has already selected humtfpb as the sequence to be annotated) to bring up a file selection menu, and use it to select the sequence file you wish to annotate. Acceptable formats for the sequence file are:
>gb|J02846|HUMTFPB Human tissue factor gene, complete cds. GAATTCTCCCAGAGGCAAACTGCCAGATGTGAGGCTGCTCTTCCTCAGTCACTATCTCTG GTCGTACCGGGCGATGCCTGAGCCAACTGACCCTCAGACCTGTGAGCCGAGCCGGTCACA CCGTGGCTGACACCGGCATTCCCACCGCCTTTCTCCTGTGCGACCCGCTAAGGGCCCCGC [etc.]
Other options that can be configured via the Genotator GUI:
~nomi/genotator/genotator -batch seq1 seq2 seq3 ...
where seq1, etc. are the names of plain or FASTA-format sequence files, and -batch tells Genotator not to bring up the GUI.
Invoking the Genotator back end with the -h option causes it to print out a list of legal
command-line options:
Usage: genotator [seqfile1 [seqfile2 ...]]
[-human or -drosophila] [-none] [-nomail] [-exit] [-d(ebug)]
[-noblast] [-nomask] [-dir output_dir] [-ace] [-exon] [-homol] [-all] [-batch]
[-grail] [-genefinder] [-genie] [-genscan] [-xpound] [-genemark]
[-genpept] [-est] [-repeats]
[-promoters] [-splice] [-trnascan] [-orf]
[-h(elp)]: print this help message
[seqfile ...]: name of sequence file in plain, FASTA, or GenBank format
(You may specify multiple sequence files.)
[-human or -drosophila]: which organism your sequence is from (human is default).
[-none]: start with no analysis boxes checked
[-nomail]: don't send email upon completion (default is to send email)
[-exit]: exit upon completion (default is not to exit)
[-d(ebug)]: debug mode (for developers)--print what Genotator would do, but don't really do it.
[-noblast]: try to reuse old BLAST output, but redo blast postprocessing
[-nomask]: don't mask out repeats before BLASTing dbEST and GenPept
[-dir output_dir]: store results in (subdirectory of) output_dir
[-exon]: run genefinders only (in batch mode)
[-all]: run all analyses in batch mode
[-batch]: run some analyses in batch mode; analyses to run will be specified by other arguments
The remaining arguments are the analyses functions that can be specified in conjunction
with the -batch option:
[-grail] [-genefinder] [-genie] [-genscan] [-xpound] [-genemark]: Run that gene finder
[-repeats]: blastn against appropriate organism's database of repeats
[-est]: blastn against EST database
[-genpept]: blastx against GenPept
[-promoters]: find promoters
[-splice]: find splice sites (as well as start/stop codons)
(When the browser is invoked, these are not displayed automatically,
but they can be turned on from the Display menu.)
[-trnascan]: look for tRNA genes
(The results aren't shown in the browser, since there virtually never are any.)
The Genotator browser can be invoked with the name of an annotated sequence file as an argument, e.g.:
~nomi/genotator/genotator-browser humtfpb
If it is invoked with no arguments, a
list of annotated sequences is displayed, with the sequences annotated
by the invoking user listed first. The other sequence directories are
collapsed and indicated by ...:
To see the sequences in a collapsed directory, double-click on the directory name (e.g. "liepe...") and the sequences (or subdirectories) in that directory will appear in the list:
If there are a lot of sequence names in the list, you can use the Find button to help you search for the one you want. When you find the sequence name you are looking for, double-click it (or single-click and hit Select). The selection list will disappear, and the browser will load the annotations for the selected sequence.
~nomi/genotator/genotator-browser -noaxis humtfpb
If you want the axis numbering to start at some number other than 0, you can specify
the start position with the -axis option, e.g.:
~nomi/genotator/genotator-browser -axis 35000 humtfpb
Be sure to specify the axis start position in bases, even though the axis is marked
in kilobases (kb). Also note that you may have to scroll slightly to the left (using the
scrollbar at the bottom of the map display) in order to see the whole number at the
left edge of the axis.
In the figure below, the Genotator browser is shown displaying the annotations on HUMTFPB, a human tissue factor gene sequence obtained from GenBank.
Each colored rectangle represents a sequence region that has been annotated. The type of annotation is identified by the color of the rectangle and also the row in which it appears. The row labels on the left (e.g. "GenPept hits") can be clicked for more information about that row. Clicking on an annotation rectangle puts a black box around the selected rectangle and displays additional information about that particular annotation in the text window at the top of the browser. This includes the start and end positions of the annotation, possibly a score, and other information. For example, if a BLAST hit is clicked, the text window might say, "BLASTX GenPept hit from 864 to 1112 with sequence gp|K01228|HUMCG1PA1_1 (33% identity)". This concise description identifies the database sequence that was hit (gp|K01228|HUMCG1PA1_1 is its GenPept ID), the region that was found to be similar to this database sequence (bases 864 to 1112), and the percentage sequence identity for the hit (33%).
BLASTX hits against GenPept can be viewed in Blixem, a BLAST hit viewer from the Sanger Centre. Blixem can be a useful tool for examining BLAST hits in more detail, but its user interface is kind of confusing.
When it comes up, Blixem shows black lines to represent hits in the region near where you clicked. The vertical position of the lines represents their percent identity. A blue box shows the region that is expanded below to show the actual hit alignments. You can move the blue box with your middle mouse button. Since BLASTX compares your DNA sequence with an amino acid database, the hits are shown in all three frames. The exact and similar matches are highlighted in color.
One of the least obvious things about Blixem is how to quit. Give up? If you click your right mouse button on some empty gray area in the Blixem display, it will pop up a menu, in which one of the choices is Quit.
Theoretically, Blixem should be able to pull up the GenBank records for subject sequences. Unfortunately, I haven't been able to get that to work; it seems to be configured specifically for the Sanger Centre. However, I recently enabled Genotator to pull up GenBank records for subject sequences; this function is available through the Display menu, and is discussed in that section.
setenv PRINTER myfavoriteprinter
One thing to keep in mind (I often forget this, so you might, too) is that if the forward strand is displayed in the sequence window, then clicking on annotations in the forward strand will highlight them in the sequence window, but clicking on annotations in the reverse strand will not. (If you are currently displaying the complement, of course, only annotations in the reverse strand will be highlighted.)
Interaction between the map and sequence displays is bidirectional: when a region is selected in the sequence display, it is boxed in the map display. Please be aware that when mousing out a region in the sequence display, if you happen to release the mouse when the cursor is midway between two rows, the boxed region in the map display will erroneously start at zero. This is a known Tkperl bug and is not fixable within Genotator.
Annotations that refer to a sizable portion of the sequence are generally added to the map; those referring to a small region (such as a primer) are more appropriately added to the sequence. All personal annotations are saved in the database along with the automatically generated annotations. Examples of personal annotations can be seen in the map display figure ("Personal annotation" and "Reverse strand annotation") and the sequence display figure ("personal annotation in sequence"). To delete personal annotations in the map display, mouse-select a box around the annotation you wish to delete (this may be easier to do if you zoom in first). The annotation will disappear from the display; however, it is not permanently gone until you hit Save Annotations. (If you quit the browser without saving, you will be asked if you want to save your personal annotations.) There is no "undo" function for deletion, but if you mistakenly delete an annotation, you can use "Reload" to reload all the personal annotations that have been saved in the database (of course, any new ones you have not yet saved will be lost).
Because personal annotations in the sequence display may overlap, the procedure for deleting them is slightly different. If you select Delete Sequence Annotation, a window will pop up showing the positions and labels on the sequence annotations. If you single-click on one of the annotations, that annotation will be highlighted in cyan on the sequence display. Double-clicking, or pressing the Delete button, will delete that annotation. The sequence display window will vanish and reappear without the deleted annotation.
/home/genotator/genotator-browser -noaxis humtfpb
If you want the axis numbering to start at some number other than 0, you can specify
the start position with the -axis option, e.g.:
/home/genotator/genotator-browser -axis 35000 humtfpb
Be sure to specify the axis start position in bases, even though the axis is marked
in kilobases (kb). Also note that you may have to scroll slightly to the left (using the
scrollbar at the bottom of the map display) in order to see the whole number at the
left edge of the axis.
@genefinders = ("GRAIL grail", "GeneFinder genefinder", "Genie genie", "GENSCAN genscan");
(The first thing in each pair is the "pretty" name by which the gene finder will be described to the user; the second is the name of the function that calls that gene finder.)
For example, if you wanted Genotator to run GeneMark instead of Genie, you could change the line to say:
@genefinders = ("GRAIL grail", "GeneFinder genefinder", "GeneMark genemark", "GENSCAN genscan");
When the browser is invoked on an annotated sequence, it dynamically
creates a row for any gene finder that was run--if it can't find
output from a given gene finder, it doesn't create a row for it. The
browser has a default order for displaying the gene finder results
(Genie, GRAIL, GeneFinder, GENSCAN, xpound, GeneMark). To change the
order, you can edit the code somewhere around line 466 of
genotator-browser.
This document was updated for Genotator version 1.21 and Genotator browser version 1.65.
Last modified:
May 3 2000