Opened 9 years ago

Closed 9 years ago

#501 closed enhancement (fixed)

bysequence or byentity keywords

Reported by: Elaine Meng Owned by: Tom Goddard
Priority: major Milestone:
Component: Command Line Version:
Keywords: Cc: Conrad Huang, Eric Pettersen, Greg Couch
Blocked By: Blocking:
Notify when closed: Platform: all
Project: ChimeraX

Description

I find the "bysequence" and "sequence" keywords (in commands "color" and "select") to be somewhat mysterious or misleading. In "color" it suggests (to me) coloring by residue type, like the NDB coloring of nucleotides and similar schemes for amino acids. In "select" it suggests I can select some sequence, e.g. "select sequence RGD" would actually select any Arg-Gly-Asp sequences in a structure.

I realize that the sequence is used to identify the same biopolymer "macromolecule entities" (as the RCSB PDB calls this concept), but suggest using keywords:

bysequence -> byentity
sequence -> entity

This would apply to "color","select" and also atomspecs if this concept were implemented as a selector as suggested in #480.

Or, if we cannot part with the bysequence/sequence keywords, could byentity/entity be added as synonyms?

Change History (3)

comment:1 by goddard@…, 9 years ago

I agree that the “sequence” options are ambiguous.  I don’t object to adding equivalent “entity” options but I think this vague term simply is not memorable, and also is ambiguous, is a ligand an entity? currently the sequence keywords only apply to polymers.  Really the sense of the “sequence” name is “sameSequence”.  Another alternative to “bysequence” would be “bypolymer”.  Basically I think “entity" is as bad as “sequence”, I’m fine having both, but would like better third option, but I have not spent the time to think of that option.

comment:2 by Elaine Meng, 9 years ago

Priority: minormajor

I have spent time thinking about it, and already used the existing keywords for several weeks to see if they would grow on me. I would not have suggested the change if I didn’t think it was a significant improvement. No single keyword can clearly convey the whole idea, so many possibilities suffer from some amount of being vague and requiring additional explanation, but I gave specific reasons why bysequence/sequence keywords are especially misleading (I'd rather get rid of them, but keeping as synonyms is the compromise). I liked that the "entity" terminology would be shared by RCSB PDB, and I don't think "polymer" "biopolymer" "macromolecule" convey the concept of sameness, whereas "color bysamesequence" and "select samesequenceas" are like smashing a sentence into one keyword and still aren't clear enough that one wouldn't need to consult the manual. Actually "entity" being a rarer word has the advantage that is less likely for people to infer wrongly what it does based on their preconceptions.

comment:3 by Tom Goddard, 9 years ago

Resolution: fixed
Status: newclosed

Uncle. I replace color command bysequence with byentity and "sequential sequences" with "sequential entities", and also select command "sequence" is now "entity". The old "sequence" options are gone.

Note: See TracTickets for help on using tickets.