BUG: binding.site() for PDB file with negative resno

[Xinqiu Yao]

binding.site() returns wrong $resno and $chain when PDB contains negative residue numbers:

> pdb <- read.pdb("4wgi")
  Note: Accessing on-line PDB file
  HEADER    APOPTOSIS/INHIBITOR                     18-SEP-14   4WGI               
   PDB has ALT records, taking A only, rm.alt=TRUE
> lig.inds <- atom.select(pdb, resid="MAL")
> prot.inds <- atom.select(pdb, "protein")
> 
> binding.site(pdb, a.inds=prot.inds, b.inds=lig.inds)
$inds

 Call:  NULL

   Atom Indices#: 175  ($atom)
   XYZ  Indices#: 525  ($xyz)

+ attr: atom, xyz

$resnames
 [1] "ASN--184 (A)" "ASP--182 (A)" "LYS--181 (A)" "GLU--152 (A)" "TRP--134 (A)"
 [6] "ALA--133 (A)" "ASP--131 (A)" "ARG--130 (A)" "GLU--85 (A)"  "ASN--46 (A)" 
[11] "GLU--43 (A)"  "PRO--42 (A)"  "TYR--41 (A)"  "PHE--40 (A)"  "TRP-34 (A)"  
[16] "MET-134 (A)"  "TRP-144 (A)"  "ARG-148 (A)" 

$resno
 [1]  NA  NA  NA  NA  NA  NA  NA  NA  NA  NA  NA  NA  NA  NA  34 134 144 148

$chain
 [1] "184" "182" "181" "152" "134" "133" "131" "130" "85"  "46"  "43"  "42" 
[13] "41"  "40"  "A"   "A"   "A"   "A"  

$call
binding.site(a = pdb, a.inds = prot.inds, b.inds = lig.inds)

[Lars Skjærven]

arrrgh

[Lars Skjærven]

> binding.site(pdb, a.inds=prot.inds, b.inds=lig.inds)
$inds

 Call:  NULL

   Atom Indices#: 175  ($atom)
   XYZ  Indices#: 525  ($xyz)

+ attr: atom, xyz

$resnames
 [1] "ASN-184 (A)" "ASP-182 (A)" "LYS-181 (A)" "GLU-152 (A)" "TRP-134 (A)"
 [6] "ALA-133 (A)" "ASP-131 (A)" "ARG-130 (A)" "GLU-85 (A)"  "ASN-46 (A)" 
[11] "GLU-43 (A)"  "PRO-42 (A)"  "TYR-41 (A)"  "PHE-40 (A)"  "TRP34 (A)"  
[16] "MET134 (A)"  "TRP144 (A)"  "ARG148 (A)" 

$resno
 [1] -184 -182 -181 -152 -134 -133 -131 -130  -85  -46  -43  -42  -41  -40   34
[16]  134  144  148

$chain
 [1] "A" "A" "A" "A" "A" "A" "A" "A" "A" "A" "A" "A" "A" "A" "A" "A" "A" "A"

$call
binding.site(a = pdb, a.inds = prot.inds, b.inds = lig.inds)

bugfix cc7d93

what do you think about the $resnames here. should there be a space there perhaps? previously I used a hyphen.

[Xinqiu Yao]

Thanks a lot!

Add space between resid and resno would be better, but I guess the $resnames is mainly for human reading (right?) and so it is fine to me with current format.

[Lars Skjærven]

Jepp.. in principle we could just omit resnames and rather make a print.bindingsite instead.