There is a good description of the Apollo tiers files in the Apollo user manual--please read that section before you read this description of how to create and change tiers files.
If you edit the tiers file, be aware that each tier line and each type line must be a single line--if these lines contain newlines, they will not be parsed properly. Some text editors will break up long lines by inserting newlines.
Note that the order of lines in the tiers file is how Apollo determines
the vertical ordering of the data in the display. The first tiers in the
file appear closest to the central axis.
Creating new feature property scheme for Gene Prediction
Creating new feature property scheme for Softberry
Creating new feature property scheme for Gene Model
Creating new feature property scheme for
blastx:Honeybee_TREMBL30.5_SWISS47.5
Creating new feature property scheme for Glean3
These messages mean that these datatypes are missing from the appropriate
tiers file and should be added to it. To figure out which tiers file Apollo
is using, look at the console output for a line like:
Using tiers file /users/nomi/.apollo/fly.tiers
The main challenges to adding these new types to the tiers file are figuring out what tier each datatype belongs in, and determining what vertical position to assign them, ranging from close to the metric to farther up in the panel.
In this example, we started with Drosophila data (which uses fly.tiers) and read a data file that contained some new types. To add these new types to fly.tiers, the curator reasoned as follows:
For Gene Prediction, this seems like some new generic gene prediction
datatype, so I add a new "Type" in the Gene Prediction tier:
[Type]
label : Gene Prediction
tiername : Gene Prediction
datatype : Gene Prediction
glyph : DrawableResultFeatureSet
color : 153,51,255
usescore : true
minscore : 100.0
column : SCORE
column : GENOMIC_RANGE
column : query_frame
column : GENOMIC_LENGTH
sortbycolumn : GENOMIC_RANGE
For Softberry, I happen to know that they make the fgenesh program, so I
add one new line with the Softberry datatype to the fgenesh Type
description:
[Type]
label : Fgenesh
tiername : Gene Prediction
datatype : fgenesh:dummy
datatype : Softberry
glyph : DrawableResultFeatureSet
color : 255,153,255
usescore : true
minscore : 100.0
column : SCORE
column : GENOMIC_RANGE
column : query_frame
column : GENOMIC_LENGTH
sortbycolumn : GENOMIC_RANGE
weburl : http://www.softberry.com/berry.phtml?topic=index&group=programs&subgroup=gfind#
I'm not sure what Gene Model is, so I create a whole new Tier
and Type for this, just after the third party annotations:
[Tier]
tiername : Gene Model
expanded : true
[Type]
label : Gene Model
tiername : Gene Model
datatype : Gene Model
glyph : DrawableResultFeatureSet
color : 255,51,51
usescore : true
minscore : 100.0
column : SCORE
column : GENOMIC_RANGE
column : MATCH_RANGE
column : GENOMIC_LENGTH
column : MATCH_LENGTH
sortbycolumn : GENOMIC_RANGE
For blastx:Honeybee_TREMBL30.5_SWISS47.5, this seems similar to BLAST
similarity to fly, but instead, it is for bee, so I created a new Tier
and Type just under the BLASTX similarity to fly with the same color:
[Tier]
tiername : BLASTX Similarity to Bee
expanded : true
[Type]
label : BLASTX Similarity to Bee
tiername : BLASTX Similarity to Bee
datatype : blastx:Honeybee_TREMBL30.5_SWISS47.5
glyph : DrawableResultFeatureSet
color : 255,20,0
usescore : true
column : SCORE
column : identity
column : expect
column : query_frame
column : GENOMIC_RANGE
column : MATCH_RANGE
column : GENOMIC_LENGTH
column : MATCH_LENGTH
sortbycolumn : GENOMIC_RANGE
groupby : HOMOLOGY
weburl :
http://srs.ebi.ac.uk/srs6bin/cgi-bin/wgetz?[swall-acc:__ID__]+-e
Types "my.tiers"
Note that you must quit and restart Apollo for your new tiers file to take effect.
Last modified: Thu Oct 27 10:14:45 PDT 2005