Hi Karin, I'm using gff3 files too, and what I've figured out so far is that by modifying the gff3.tiers file you can get Apollo to display the characteristics you want. Specifically, in the [type] definition the field "datatype" corresponds to the 3rd column, and "resulttype" corresponds to the 2nd column. This is useful if you want to write out by hand new [type] definitions, but that's not necessary. Also, if you have specific information you'd like to be displayed that doesn't correspond to any of the columns in gff3, you can add any number of tags in the 9th column. A tag is of the form name=value, and they are separated by ; You can have these tag values displayed in the right-hand box by adding the line column : tag_name to the type definition in the gff3.tiers file.
concretely speaking, if you load a fasta file with a sequence called scaffold00028 and a gff3 file that has lines that look like this: ##gff-version 3 scaffold00028 blastn:CuMe_transp match 102186 107159 0.0 + 0 Target_seq=gi|148270933:37317-42727;per_id=89.7426859358477;desc=Cucumis melo clone BAC 1-21-10 retrotranspo Ty3 gypsy; you'll see the results appear in the black portion of the screen. For this, it is necessary that the 3rd column be "match", which defines the results as computational results, and not annotations (which belong in the blue part of the screen) if you open the types window, you'll see that a new type has appeared, in this case "blastn:CuMe_transp", which is the source, or 2nd column. then, go to edit->preferences and click on the "types" tab in the tiers drop-down menu, select your new tier (in our case, blastn:CuMe_transp) change the color, or the shape, or something and press "save". the idea is to change something, so that when you save it there will be a new entry written into the gff3.tiers file. actually, what happens is that a new gff3.tiers file is written to the .apollo directory with the new entry. The default one is in the usr/local/Apollo/conf/ directory, and if you make any modifications to a tier a new one is created and put in your home/your_name/.apollo directory, and that's the one that's used. The .apollo directory is hidden, so be sure to turn on "view hidden files" in your browser to see it. (this is for linux, I'm not quite sure where it'd be on a windows machine or a mac) so now, if you open the new gff3.tiers file, you'll see that there's an entry that looks like this: [Tier] tiername : BLASTN:CuMe_transp [Type] label : BLASTN:CuMe_transp tiername : BLASTN:CuMe_transp datatype : BLASTN:CuMe_transp glyph : DrawableGeneFeatureSet color : yellow utr_color : yellow column : GENOMIC_LENGTH 1 column : GENOMIC_RANGE 1 column : SCORE 1 column : NAME 1 to see additional feature information in the bottom-right box, you can add any number of "column : tag_name" lines. These will refer to the types of tags in the 9th column. So, if I want to see displayed the values of all the tags I have in the 9th column, I'd add column : evalue column : Target_seq column : desc column : per_id and save it. You have to re-start apollo for it to implement the changes you just made. Well, I hope that helps, and sorry if any of this is something you already knew :) Cheers, Elizabeth Ed Lee <e...@...> writes: > > Hi Karin, > > I'd recommend using GFF3 as it's actively supported. You can get > information on the GFF3 spec at: > > http://www.sequenceontology.org/gff3.shtml > > If you have issues with the GFF3 files, please let us know. > > Cheers, > Ed > > On Mon, 10 May 2010, Karin Lagesen wrote: > > > Hi! > > > > I am trying to figure out what goes where when I load a gff3 or ensembl > > gff formated file. I see that the information from those files are > > displayed in the bottom-most windows (as well as being used to draw the > > genes, of course). However, I am uncertain where the information from each > > column goes, and whether this varies according to what the source is. For > > instance, in a gff ensembl formatted file I have (amongst others) the > > following lines: > > > > STAT_bind1 GeneWise cds 232019 232107 0.00 + 0 > > GeneScaffold_1701-genewise-prediction-1 > > STAT_bind1 GeneWise cds 232219 232279 0.00 + 1 > > GeneScaffold_1701-genewise-prediction-1 > > STAT_bind2 GeneWise cds 232538 232633 0.00 + 0 > > GeneScaffold_1701-genewise-prediction-2 > > STAT_bind2 GeneWise cds 232732 232779 0.00 + 0 > > GeneScaffold_1701-genewise-prediction-2 > > STAT_bind2 GeneWise cds 235325 235408 0.00 + 0 > > GeneScaffold_1701-genewise-prediction-2 > > 90107 WGA2Genes similarity 167067 167276 0.0 + . > > ENSGACP00000011294 41 110 > > 90112 WGA2Genes similarity 173628 173771 0.0 + . > > ENSGACP00000011294 112 159 > > 90116 WGA2Genes similarity 174150 174275 0.0 + . > > ENSGACP00000011294 160 201 > > 75089 combined__ensembl exon 91409 91415 0.0 + . > > 5772 > > 75092 combined__ensembl exon 91421 91614 0.0 + . > > 5772 > > 75095 combined__ensembl exon 92106 92196 0.0 + . > > 5772 > > 75097 combined__ensembl exon 92362 92495 0.0 + . > > 5772 > > 75099 combined__ensembl exon 92735 92798 0.0 + . > > 5772 > > 75101 combined__ensembl exon 92917 93104 0.0 + . > > 5772 > > > > > > So, here I have cdses, exons and similarities. Now, when I look at these > > in apollo and click on one of those that are from GeneWise, I get as a > > heading in the bottom right window STAT_bind2. A similar thing happens for > > the combined_ensembl genes, where the heading for some reason becomes one > > of the rightmost columns for the combined__ensembl. This does not happen > > for the WGA2Genes, where I don't get the same kind of heading in the > > bottom right window. In this case I just get WGA2Genes. > > > > So, basically, is there any where I can get a detailed breakdown of what > > the information that is found in a gff3 and/or ensembl gff formatted file > > can be refound/used for in apollo? > > > > Thanks! > > > > Karin > > -- > > Karin Lagesen, PhD > > > > > > _______________________________________________ > > apollo mailing list > > apo...@... > > http://mail.fruitfly.org/mailman/listinfo/apollo > > > > >
_______________________________________________ apollo mailing list [email protected] http://mail.fruitfly.org/mailman/listinfo/apollo
