IUBio GIL .. BIOSCI/Bionet News .. Biosequences .. Software .. FTP

How is PEPMAP integrated into the rest of the acedb?

Richard Durbin rd at sanger.ac.uk
Wed Jun 23 06:31:16 EST 1999


Hello Mark, Doug and others,

PEPMAP is a specialised display module in acedb, comparable to the
FMAP (the display module for DNA sequence) or GMAP (the display module
for genetic and other types of maps).  I am sorry there is not much
documentation.  It was not developed by Benedict Arnold, although he
may have used it, but rather by a succession of people at the Sanger
Centre including Erik Sonnhammer, Clive Brown and Ian Longden.

The key sections of models that it uses are the following:

?Protein  Title UNIQUE ?Text
	  Peptide UNIQUE ?Peptide UNIQUE Int UNIQUE Int // Int Int are length, checksum
	  View_info View ?View
	            Default_view UNIQUE ?View
	  Homol	DNA_homol ?Sequence XREF Pep_homol ?Method Float Int UNIQUE Int Int UNIQUE Int #Homol_info
		Pep_homol ?Protein XREF Pep_homol ?Method Float Int UNIQUE Int Int UNIQUE Int #Homol_info
		Motif_homol ?Motif XREF Pep_homol ?Method Float Int UNIQUE Int Int UNIQUE Int #Homol_info
	    // Float Int Int Int Int are score, start end in protein, start end in target
	  Feature ?Method Int Int UNIQUE Float	// display according to method
	    // Int Int Float are start, end, score

with reciprocal entries in ?Sequence, ?Motif for homology information.

?Peptide is an array class.  You should enter protein sequence as
follows:

Peptide WP:K04C2.3
MATQTAVTATTVPQGSPSVTGSASQGGILNSARQFFSLAAPKVEKDKEFSFRFLTSP
LRSKPAATVNADNFYMYACGSINMP

similar to the way DNA is entered.  This will automatically create the
Protein class object with the same name, and enter the link to the
peptide with the length and checksum.

The main way to enter annotation for the protein sequence is via Homol
and Feature lines, exactly as for DNA.  The way they are displayed are
controlled by the Method object, with a different method for each
different type of Match/Feature.  The relevant parts of the ?Method
model are (I think)

?Method	Remark ?Text
	  // the Display information controls how the column looks.
	Display Colour #Colour
		Score	Score_by_offset	// has priority over width, for Jean
			Score_by_width
			Score_bounds UNIQUE Float UNIQUE Float 
				// limits for squashing
		Bumpable	  // bump to avoid overlap
	Blixem_P
	Belvu

The PEPMAP display uses the same system of columns controlled by a
View object as the GMAP display.  The relevant parts of the View
models are given below, but although you need them you don't really
need to understand them - they are all handled by the code and you
should not enter data for them directly.

?View 	Type 	UNIQUE 	Pepmap
	Display Submenus	// BOOL for submenus on item boxes
		No_buttons      // suppress header buttons for WWW 
		Hide_header     // No headers and footers 
	Name 	UNIQUE 	Text
	Columns	Text UNIQUE Int #Column	// Int is BOOL for Hidden/Visible

?Column	UNIQUE	Scale Scale_unit UNIQUE Float	// minimum increment
		      Cursor Cursor_on
			     Cursor_unit UNIQUE Float
		Locator Magnification UNIQUE Float
			Projection_lines_on
		pepSequence	PS_Highlight_residue   // il all needed for colomn display of peptides
				PS_Residues_per_wrap Int
				PS_Colours Text #Colour
		Hydrophobicity	HP_Show_Zero_bar
				HP_Fixed_Scaling
				HP_Calculation_window Int
				HP_Display_width Int
                Homol           HOM_bump
                pepFeature      FEA_bump
                                FEA_Query Text
                Homol_Name      HOM_NAME_bump
				HOM_NAME_width Int
                pepActiveZone
	        

there are other column types and parts of the ?View module needed for
other displays.  The list of columns shows what sort of information
can be shown in the PEPMAP.  A default View is constructed
automatically, but quite a lot is configurable, and configurations can
be stored as named View objects, and proteins can have default and
alternative views.  e.g. the sequence itself can be displayed with
different colouring schemes, configurable from the view control window
obtained by pressing the "View" button on the peptide display.

It looks like by default there are columns for BlastP, BlastX,
Pfam-hmmls and Pfam-hmmfs.  Anything other methods need adding
explicitly to the View in the View control window.  You can display
multiple methods in one column by separating the method names with
'+', e.g. "Pfam-hmmls+Pfam-hmmfs".  This is all a bit hard-wired for
our priorities, sorry.

Also, the PEPMAP has not been continuously developed and upgraded over
the last few years, so is lacking gif drawing support for gifaceserver
and new Method/Column options available in FMAP.  But it is functional
and quite powerful/flexible for displaying new sorts of data.

Richard

PS Can someone at the Sanger Centre turn this email into a
whelp/Protein-display.html document and add "-help Protein-display" to
the wspec/displays.wrm file, so there is documentation for the
display?
---




More information about the Acedb mailing list

Send comments to us at archive@iubioarchive.bio.net