Peter, Carton, some of this functionality is already present in Artemis. If you load up a codon preference table before marking all reading frames, then each ORF will be given a /score qualifier from its match to the table, and will be coloured from pale to dark blue depending on the score value.
In addition, using the right hand mouse menu, you can filter the visible features using "set score cutoffs" (this only works on the features on screen). You can very rapidly filter out junk by zooming out until the whole sequence is visible, using "set score cutoff" to see only the good scores, then "select visible features". You can either copy those figures to a new entry or close the score cutoffs filter, use "toggle selection" and delete the junk. Of course, this is only a crude tool, and is not a substitute for using a proper gene prediction program. As Peter says, it is possible to add /score and /colour=R G B to features before loading, and use a similar system. The problem with scoring or colouring ORFs acording to GC3 values is that the meaning of the GC3 is different depending on the GC content of the sequence as a whole - it would be very difficult to make Artemis behave consistently with different organisms. I'm not sure of the original reason for marking ORFs stop-to-stop; much of it is historical. It is more useful for spliced genomes, and for some specific cases. It should be easy to add an extra menu item to mark start-to-stop as well, and this could be a useful addition. For ACT, you might notice that the new version colours forward matches red and reverse matches blue (still with a light-dark scale), in order to make zoomed-out views easier to interpret. We have discussed being able to load and toggle different comparison files, but it hasn't made it to the top of the priority list. Julian. Peter Young wrote: > > Hi Kim and everyone > > I would echo Carton's idea that it would be good to have a way to colour > each CDS to reflect GC3, but I would extend it further because there are > potentially other quantitative values that one could associate with > genes and display visually in Artemis. What about gene expression data, > for example? Or Blast scores? Of course, these cannot be calculated > from the sequence (unlike GC3), so would need the user to provide a file > with values for each feature that Artemis would then translate into > colours and display. > > Actually, I think it should already be possible to do this, because > Artemis understands a "/colour=R G B" qualifier. So if Carton writes a > script to calculate GC3, convert this into RGB colour values and insert > the appropriate /colour qualifier for each CDS into the file, he can see > his genes in glorious colour. It would be nice if the Artemis people > made it easier, though! > > On a related note, we would like to be able to use different colours in > the comparison window of ACT. As well as the current red and pinks for > the lines connecting matches, we would like to use other colours to show > different classes of match (in our case, genes supporting different > phylogenies, but it could be functional classes, or whatever). How > about allowing multiple comparison files to be loaded, with the user > choosing the colour for each? > > Peter > > Kim Rutherford wrote: > > > Hi Carton. > > > > On Mon, 23 Aug 2004 21:15:41 +0800, Carton Chen wrote: > > > > > >>Dear Kim, > >>I saw the new version of Artemis. I downloaded it, add the ,command > >>suffix and test ran it a little bit. I like it. It is significantly > >>faster. more responsive, and more convenient. > > > > > >>A thought occurred to me recently regarding Artemis. It would be great > >>if you let the ORF arrows show info regarding their codon preferences > >>in color: Different colors (say from Red to orange to yellow to green > >>to blue to cyan) for different range (100 to 0%) of G/C in the third > >>position of codon, This way one may look at the ORF and readily decide > >>which one is likely coding in G+C rich or G+C poor genomes. > > > > > > That's an interesting suggestion. I'm CCing this to the Artemis > > mailing list so others can comment. > > > > > > > >>Another alternative of course is to give the ORF the same color that > >>matches the corresponding curve in the CC Frame Plot. That would be > >>very helpful, too. > > > > > > That sort of thing has been discussed in the past, but it never made > > it to the top of the priority list. > > > > > > > >>Another suggestion: > >> At least in prokaryote mode (if there is), the initial ORF mapped > >>should start with the first potential initiation codon after the last > >>stop codon, not just the first codon after. The latter adds a lot extra > >>work - at least one initial move for each potential ORF. > > > > > > I'm sure there is reason for this behaviour. I'll leave it to others > > to comment. > > > > Kim. > > -- > Prof J P W Young > Department of Biology 3 > University of York > P O Box 373 > York YO10 5YW, UK > > [EMAIL PROTECTED] > tel: +44 1904 328630(direct), 328500(department) > fax: +44 1904 328505(department) > http://www.york.ac.uk/depts/biol/staff/jpwy/jpwy.htm > == > -- Julian Parkhill Senior Investigator, The Sanger Institute, Wellcome Trust Genome Campus, Hinxton, Cambridge, CB10 1SA, UK Visiting Professor in Microbial Genomics, University of Oxford Tel: +44(0)1223 494975, Fax: +44(0)1223 494919 http://www.sanger.ac.uk/Projects/Microbes http://www.sanger.ac.uk/Teams/faculty/parkhill
