Menu
▾
▴
[Gmod-apollo-cmts] CVS: apollo/doc todo,1.253,1.254
|
From: Nomi H. <no...@us...> - 2003-04-15 18:24:24
|
Update of /cvsroot/gmod/apollo/doc In directory sc8-pr-cvs1:/tmp/cvs-serv31826 Modified Files: todo Log Message: added new section for synteny view bugs Index: todo =================================================================== RCS file: /cvsroot/gmod/apollo/doc/todo,v retrieving revision 1.253 retrieving revision 1.254 diff -C2 -d -r1.253 -r1.254 *** todo 14 Apr 2003 21:54:58 -0000 1.253 --- todo 15 Apr 2003 18:24:15 -0000 1.254 *************** *** 373,450 **** the white text. - -------------------- - GFF/Ensembl problems - -------------------- - - Need to be more robust and informative when user makes bad GFF or Ensembl - CGI requests: - - If you ask for a gff file that's not found, you get a popup that just - says Exception from apollo.dataadapter.DataAdapterException. Message - "Error: could not open file foo for reading." is just printed to - terminal. - - Ensembl CGI data adapter doesn't deal gracefully with bad ranges. For example, if you - ask for a range starting at 0, you get many exceptions. It should pop up a box saying - "The valid range for chromosome N is 1-3592309" or whatever. - - If you put letters instead of numbers in the boxes, it appears to freeze up. - - Ensj/Otter also behaves badly if you request nonexistent chromosomes or - bad ranges - - - Restriction enzyme finder doesn't work on Ensembl data, even when - you're zoomed in enough to see all the bases! Also get message - "Warning: No features to getEnd() for". - - A lot of Ensembl results, when you click on them, show names like - "sptrembl: no_name" in the righthand detail panel. If second part of - name is "no_name", don't show it. But there are some results that seem - like they should have real names. - - Got rid of the no_name by changing SetDetailPanel:featureDisplayString, - but line below colored line on right has nothing in it--should have - something. - - GFF parsing problems: - - When fixing the no_name problem, I put a println in DefaultNameAdapter:getDisplayName - and discovered that this method is called whenever you move the mouse over the table - in the left detail panel! This isn't terrible, but it seems like a waste of time. - I'm not sure why it's getting called so much--I couldn't track down the event that - caused it. - - As an experiment, I tried opening data/example.xml and then using Save As - to save it in GFF format. When I then tried to load the resulting GFF - file, I got an exception. When I fixed the GFF writer to take care of that - exception, it uncovered a whole bunch of problems--I think the GAME->GFF - conversion is not working right at all. Ack! - - Saving a GFF file in GAME format also doesn't work. I fixed some problems, - but we're getting ClassCastExceptions. - - Problems starting up Apollo with GFF file + FASTA file (reported by Kim - Worley): - /bin/apollo -f ./data/HAWK/hawk.anno.gff hawk ./data/stitch.chr9.fold.fa.masked - java.lang.NullPointerException: Can't accept feature name of null. - Use RangeI.NO_NAME instead. - at apollo.datamodel.Range.setName(Range.java:23) - at apollo.dataadapter.GFFAdapter.getCurationSet(GFFAdapter.java:83) - - Can't load gff + fasta (see bug reports in bug tracker, - http://sourceforge.net/tracker/?atid=462763&group_id=27707&func=browse) - - Ensj bugs: - - Alex reported that stop and start dont adjust when moving splice - sites for ensj fugu. I found this to be true in ensj but couldnt - repeat in game. - - When you set ensj to fugu db a popup comes up saying it cant find - the db - but in fact it does. - - ensj with annot view showing a popup comes up saying "No valid - annotation server selected...", but things load ok. - - When looking at Ensembl data, tried to drag a stop codon, got exception: - HEY!, couldn't create a apollo.gui.drawable.DrawableTerminalCodon for stopcodon_frame2 because java.lang.ClassNotFoundException: apollo.gui.drawable.DrawableTerminalCodon - java.lang.ClassNotFoundException: apollo.gui.drawable.DrawableTerminalCodon - at java.net.URLClassLoader$1.run(URLClassLoader.java:200) - ... - at java.lang.Class.forName(Class.java:120) - at apollo.gui.drawable.DrawableFeatureSet.createDrawableFeature(DrawableFeatureSet.java:150) - ... - - --------------------------- - End of GFF/Ensembl problems - --------------------------- -- --- 373,376 ---- *************** *** 652,659 **** current region), shouldn't try to go to it in the EDE. - -------------------- ! Desired new features -------------------- Put in some sort of marker where sequence/region ends. Could be --- 578,690 ---- current region), shouldn't try to go to it in the EDE. -------------------- ! GFF/Ensembl problems -------------------- + + Need to be more robust and informative when user makes bad GFF or Ensembl + CGI requests: + - If you ask for a gff file that's not found, you get a popup that just + says Exception from apollo.dataadapter.DataAdapterException. Message + "Error: could not open file foo for reading." is just printed to + terminal. + - Ensembl CGI data adapter doesn't deal gracefully with bad ranges. For example, if you + ask for a range starting at 0, you get many exceptions. It should pop up a box saying + "The valid range for chromosome N is 1-3592309" or whatever. + - If you put letters instead of numbers in the boxes, it appears to freeze up. + - Ensj/Otter also behaves badly if you request nonexistent chromosomes or + bad ranges + + - Restriction enzyme finder doesn't work on Ensembl data, even when + you're zoomed in enough to see all the bases! Also get message + "Warning: No features to getEnd() for". + + A lot of Ensembl results, when you click on them, show names like + "sptrembl: no_name" in the righthand detail panel. If second part of + name is "no_name", don't show it. But there are some results that seem + like they should have real names. + - Got rid of the no_name by changing SetDetailPanel:featureDisplayString, + but line below colored line on right has nothing in it--should have + something. + + GFF parsing problems: + - When fixing the no_name problem, I put a println in DefaultNameAdapter:getDisplayName + and discovered that this method is called whenever you move the mouse over the table + in the left detail panel! This isn't terrible, but it seems like a waste of time. + I'm not sure why it's getting called so much--I couldn't track down the event that + caused it. + - As an experiment, I tried opening data/example.xml and then using Save As + to save it in GFF format. When I then tried to load the resulting GFF + file, I got an exception. When I fixed the GFF writer to take care of that + exception, it uncovered a whole bunch of problems--I think the GAME->GFF + conversion is not working right at all. Ack! + - Saving a GFF file in GAME format also doesn't work. I fixed some problems, + but we're getting ClassCastExceptions. + - Problems starting up Apollo with GFF file + FASTA file (reported by Kim + Worley): + /bin/apollo -f ./data/HAWK/hawk.anno.gff hawk ./data/stitch.chr9.fold.fa.masked + java.lang.NullPointerException: Can't accept feature name of null. + Use RangeI.NO_NAME instead. + at apollo.datamodel.Range.setName(Range.java:23) + at apollo.dataadapter.GFFAdapter.getCurationSet(GFFAdapter.java:83) + + Can't load gff + fasta (see bug reports in bug tracker, + http://sourceforge.net/tracker/?atid=462763&group_id=27707&func=browse) + + Ensj bugs: + - Alex reported that stop and start dont adjust when moving splice + sites for ensj fugu. I found this to be true in ensj but couldnt + repeat in game. + - When you set ensj to fugu db a popup comes up saying it cant find + the db - but in fact it does. + - ensj with annot view showing a popup comes up saying "No valid + annotation server selected...", but things load ok. + + When looking at Ensembl data, tried to drag a stop codon, got exception: + HEY!, couldn't create a apollo.gui.drawable.DrawableTerminalCodon for stopcodon_frame2 because java.lang.ClassNotFoundException: apollo.gui.drawable.DrawableTerminalCodon + java.lang.ClassNotFoundException: apollo.gui.drawable.DrawableTerminalCodon + at java.net.URLClassLoader$1.run(URLClassLoader.java:200) + ... + at java.lang.Class.forName(Class.java:120) + at apollo.gui.drawable.DrawableFeatureSet.createDrawableFeature(DrawableFeatureSet.java:150) + + --------------------------- + End of GFF/Ensembl problems + --------------------------- + + --------------------------- + Synteny view problems + --------------------------- + + After loading GFF synteny files, I pulled up the data adapter chooser and + pulled down the list, and now all that's on it is Synteny! + + Need to preload defaults so user can see a synteny demo w/o going through + a bunch of steps. + + History for Synteny GFF doesn't seem to be working. + + Can't ask user to keep changing names of synteny files. Need a better + way to do that. + + The "Species1", "Species2-Species1", "Species2" File menu items don't + appear to do anything. They should be checkboxes. Also, they should be + in the View menu, not the File menu--or, even better, there should be a + new Synteny menu. + + When you select "Species2-Species1", options in the View menu still seem + to apply only to one panel, not both. + + Many menu items (e.g. Make evidence panel vertical) throw exceptions. + Probably need to disable some menu items when in synteny mode. + + ---------------------------- + End of synteny view problems + ---------------------------- + + + + ------------------------------------------------------------------- + Desired new features + ------------------------------------------------------------------- Put in some sort of marker where sequence/region ends. Could be |