Code
--[[
SS Edit
Get and set secondary structure
SS Edit displays the current secondary structure.
The displayed value can be selected and cut or copied.
A new value can be pasted in. When the "Change" button
is clicked, the currently displayed secondary structure
will be applied to the protein.
SS Edit doesn't validate or edit the secondary structure
codes. Only the foldit codes "H" for helix, "E" for sheet,
and "L" for loop should be used. Other values may produce
unpredictable results.
If the structure list is longer than the protein, SS Edit
discards the extra entries at the end of the list.
If the structure list is shorter than the protein, SS Edit
applies the list to the first *n* segments of the protein,
where *n* is the length of the list. Any remaining segments
are unchanged.
All changes are written to the scriptlog.
See "AA Copy Paste Compare v 1.1.1 -- Brow42" for
a full-function recipe that works with primary and
secondary structures.
version 1.1 -- 2015/07/06 -- LociOiling
* speed things up a bit by disabling filters
* add standard cleanup
* don't crash trying to set type "M", skip invalid codes
version 1.2 -- 2016/12/23 -- LociOiling
* a string is not a table
* enable 1-step undo with undo.SetUndo ( false )
version 2.0 -- 2020/03/10 -- LociOiling
* add chain awareness to match AA Edit 2.0
* copy chain logic from AA Edit 2.0, expand to get SS
version 2.0.1 -- 2020/04/16 -- LociOiling
* handle proline at N-terminal correctly
]]--
--
-- Globals
--
Recipe = "SS Edit"
Version = "2.0.1"
ReVersion = Recipe .. " v." .. Version
STYPES = {
E = { "sheet" },
H = { "helix" },
L = { "loop" },
}
AALONG = 1
AACODE = 2 -- redundant for proteins, needed for DNA and RNA
AAATOM = 3
AATYPE = 4
--
-- amino acid names and abbeviations,
-- third element is mid-chain atom count
--
AANames = {
a = { "alanine", "a", 10, "P", },
c = { "cysteine", "c", 11, "P", },
d = { "aspartate", "d", 12, "P", },
e = { "glutamate", "e", 15, "P", },
f = { "phenylalanine", "f", 20, "P", },
g = { "glycine", "g", 7, "P", },
h = { "histidine", "h", 17, "P", },
i = { "isoleucine", "i", 19, "P", },
k = { "lysine", "k", 22, "P", },
l = { "leucine", "l", 19, "P", },
m = { "methionine", "m", 17, "P", },
n = { "asparagine", "n", 14, "P", },
p = { "proline", "p", 15, "P", },
q = { "glutamine", "q", 17, "P", },
r = { "arginine", "r", 24, "P", },
s = { "serine", "s", 11, "P", },
t = { "threonine", "t", 14, "P", },
v = { "valine", "v", 16, "P", },
w = { "tryptophan", "w", 24, "P", },
y = { "tyrosine", "y", 21, "P", },
--
-- bonus! codes for ligands ("x" is common, but "unk" is historic)
--
x = { "ligand", "x", 0, "M", },
unk = { "ligand", "x", 0, "M", },
--
-- bonus! RNA nucleotides
--
ra = { "adenine", "a", 0, "R", },
rc = { "cytosine", "c", 0, "R", },
rg = { "guanine", "g", 0, "R", },
ru = { "uracil", "u", 0, "R", },
--
-- bonus! DNA nucleotides (as seen in PDB, not confirmed for Foldit)
--
da = { "adenine", "a", 0, "D", },
dc = { "cytosine", "c", 0, "D", },
dg = { "guanine", "g", 0, "D", },
dt = { "thymine", "t", 0, "D", },
}
--
-- SSNames is to parallel AANames,
-- but not as complex
--
SSNames = {
H = { "helix", },
E = { "sheet", },
L = { "loop", },
M = { "ligand", },
}
AA_ATOM_MAX = 27 -- modified AA if over this count
--
-- tables for converting external nucleobase codes to Foldit internal codes
--
RNAin = {
a = "ra",
c = "rc",
g = "rg",
u = "ru",
}
DNAin = {
a = "da",
c = "dc",
g = "dg",
t = "dt",
}
Ctypes = {
P = "protein",
D = "DNA",
R = "RNA",
M = "ligand",
}
--
-- begin protNfo Beta package version 0.2a
--
-- version 0.2a is packaged as a psuedo-class or psuedo-module
-- containing a mix of data fields and functions
--
-- all entries must be terminated with a comma to keep Lua happy
--
-- the commas aren't necessary if only function definitions are present
--
-- removed some items found in 0.1 not needed here,
-- added N-terminal and C-terminal checks, first and last analysis
--
-- this version depends on the external AANames table and associated codes,
-- so still a work in progress
--
-- version 0.2a contains a quick fix for proline at N-terminal
--
-- need to reconcile this version with the more extensive version in print protein
--
protNfo = {
PROTEIN = "P",
LIGAND = "M",
RNA = "R",
DNA = "D",
UNKNOWN_AA = "x",
UNKNOWN_BASE = "xx",
CYSTEINE_AA = "c",
PROLINE_AA = "p",
aa = {}, -- amino acid codes
ss = {}, -- secondary structure codes
atom = {}, -- atom counts
mute = {}, -- mutable flag
ctype = {}, -- segment type - P, M, R, D
first = {}, -- true if segment is first in chain
last = {}, -- true if segment is last in chain
nterm = {}, -- true if protein and if n-terminal
cterm = {}, -- true if protein and if c-terminal
fastac = {}, -- external code for FASTA-style output
setNfo = function ()
local segCnt = structure.GetCount ()
--
-- initial scan: retrieve basic information from Foldit
--
for ii = 1, segCnt do
local nterm = false
local cterm = false
protNfo.aa [ #protNfo.aa + 1 ] = structure.GetAminoAcid ( ii )
protNfo.ss [ #protNfo.ss + 1 ] = structure.GetSecondaryStructure ( ii )
protNfo.atom [ #protNfo.atom + 1 ] = structure.GetAtomCount ( ii )
protNfo.mute [ #protNfo.mute + 1 ] = structure.IsMutable ( ii )
local aatab = AANames [ protNfo.aa [ ii ] ]
if aatab ~= nil then
protNfo.ctype [ #protNfo.ctype + 1 ] = aatab [ AATYPE ]
--
-- special case for puzzles 879, 1378b, and similar
--
-- if unknown amino acid, but secondary structure is not
-- ligand, mark it as protein
--
-- segment 134 in puzzle 879 is the example
--
if protNfo.ctype [ ii ] == protNfo.LIGAND
and protNfo.ss [ ii ] ~= protNfo.LIGAND then
protNfo.ctype [ ii ] = protNfo.PROTEIN
end
else
protNfo.ctype [ #protNfo.ctype + 1 ] = protNfo.LIGAND
aa = protNfo.UNKNOWN_AA
end
--
-- for proteins, determine n-terminal and c-terminal
-- based on atom count
--
if protNfo.ctype [ ii ] == protNfo.PROTEIN then
local ttyp = ""
local noteable = false
local ac = protNfo.atom [ ii ] -- actual atom count
local act = aatab [ AAATOM ] -- reference mid-chain atom count
if ac ~= act
or ( protNfo.aa [ ii ] == protNfo.CYSTEINE_AA and ac == act ) then
ttyp = "non-standard amino acid"
if ac == act + 2 then
ttyp = "N-terminal"
nterm = true
notable = true
elseif ac == act + 1 then
ttyp = "C-terminal"
cterm = true
notable = true
elseif protNfo.aa [ ii ] == protNfo.PROLINE_AA and ac == act + 3 then
ttyp = "N-terminal"
nterm = true
notable = true
end
if protNfo.aa [ ii ] == protNfo.CYSTEINE_AA then
local ds = current.GetSegmentEnergySubscore ( ii, "Disulfides" )
-- print ( "cysteine at " .. ii .. ", disulfides score = " .. ds )
if ds ~= 0 and math.abs ( ds ) > 0.01 then
nterm = false
cterm = false
ttyp = "disulfide bridge"
if ac == act + 1 then
ttyp = "N-terminal"
nterm = true
elseif ac == act then
ttyp = "C-terminal"
cterm = true
end
notable = true
else
ttyp = "unpaired cysteine"
notable = false
end
end
if notable then
print ( ttyp ..
" detected at segment "
.. ii ..
", amino acid = \'"
.. protNfo.aa [ ii ] ..
"\', atom count = "
.. ac ..
", reference count = "
.. act ..
", secondary structure = "
.. protNfo.ss [ ii ]
)
end
end
end
if protNfo.ctype [ ii ] == protNfo.LIGAND then
print ( "ligand detected at segment " .. ii )
end
protNfo.nterm [ #protNfo.nterm + 1 ] = nterm
protNfo.cterm [ #protNfo.cterm + 1 ] = cterm
protNfo.fastac [ #protNfo.fastac + 1 ] = aatab [ AACODE ]
end
--
-- rescan to determine first and last in chain for all types
-- it's necessary to "peek" at neighbors for DNA and RNA
--
for ii = 1, segCnt do
local nterm = protNfo.nterm [ ii ]
local cterm = protNfo.cterm [ ii ]
local first = false
local last = false
if ii == 1 then
first = true
end
if ii == segCnt then
last = true
end
if protNfo.ctype [ ii ] == protNfo.PROTEIN then
if protNfo.nterm [ ii ] then
first = true
end
if protNfo.cterm [ ii ] then
last = true
end
--
-- special case for puzzles 879, 1378b, and similar
--
-- if modified AA ends or begins a chain, mark
-- it as C-terminal or N-terminal
--
-- hypothetical: no way to test so far!
--
if AANames [ protNfo.aa [ ii ] ] [ AACODE ] == protNfo.UNKNOWN_AA then
if ii > 1 and protNfo.ctype [ ii - 1 ] ~= protNfo.ctype [ ii ] then
first = true
protNfo.nterm [ ii ] = true
print ( "non-standard amino acid at segment "
.. ii ..
" marked as N-terminal" )
end
if ii < segCnt and protNfo.ctype [ ii + 1 ] ~= protNfo.ctype [ ii ] then
last = true
protNfo.cterm [ ii ] = true
print ( "non-standard amino acid at segment "
.. ii ..
" marked as C-terminal" )
end
end
elseif protNfo.ctype [ ii ] == protNfo.DNA
or protNfo.ctype [ ii ] == protNfo.RNA then
if ii > 1 and protNfo.ctype [ ii - 1 ] ~= protNfo.ctype [ ii ] then
first = true
end
if ii < segCnt and protNfo.ctype [ ii + 1 ] ~= protNfo.ctype [ ii ] then
last = true
end
else -- ligand
first = true
last = true
end
protNfo.first [ #protNfo.first + 1 ] = first
protNfo.last [ #protNfo.last + 1 ] = last
end
end,
}
--
-- end protNfo Beta package version 0.2
--
--
-- end of globals section
--
function getChains ()
--
-- getChains - build a table of the chains found
--
-- Most Foldit puzzles contain only a single protein (peptide) chain.
-- A few puzzles contain ligands, and some puzzles have had two
-- protein chains. Foldit puzzles may also contain RNA or DNA.
--
-- For proteins, the atom count can be used to identify the first
-- (N terminal) and last (C terminal) ends of the chain. The AANames
-- table has the mid-chain atom counts for each amino acid.
--
-- Cysteine is a special case, since the presence of a disulfide
-- bridge also changes the atom count.
--
-- For DNA and RNA, the beginning and end of the chain is determined
-- by context at present. For example, if the previous segment was protein
-- and this segment is DNA, it's the start of a chain.
--
-- Each ligand is treated as a chain of its own, with a length of 1.
--
-- chain table entries
-- -------------------
--
-- ctype - chain type - "P" for protein, "M" for ligand, "R" for RNA, "D" for DNA
-- fasta - FASTA-format sequence, single-letter codes (does not include FASTA header)
-- fastab - "backup" of fasta
-- sstruc - secondary structure sequence, parallel to fasta, H/E/L
-- sstrucb - "backup" of sstruc, parallel to fastab
-- start - Foldit segment number of sequence start
-- stop - Foldit segment number of sequence end
-- len - length of sequence
-- chainid - chain id assigned to entry, "A", "B", "C", and so on
--
-- For DNA and RNA, fasta and fastab contain single-letter codes, so "a" for adenine.
-- The codes overlap the amino acid codes (for example, "a" for alanine).
-- The DNA and RNA codes must be converted to the appropriate two-letter codes Foldit
-- uses internally, for example "ra" for RNA adenine and "da" for DNA adenine.
--
--
-- we're assuming Foldit won't ever have more chains
--
local chainid = { "A", "B", "C", "D", "E", "F", "G", "H", "I", "J", "K", "L", "M",
"N", "O", "P", "Q", "R", "S", "T", "U", "V", "W", "X", "Y", "Z" }
local chainz = {}
local chindx = 0
local curchn = nil
local segCnt = structure.GetCount ()
for ii = 1, segCnt do
if protNfo.first [ ii ] then
chindx = chindx + 1
chainz [ chindx ] = {}
curchn = chainz [ chindx ]
curchn.ctype = protNfo.ctype [ ii ]
curchn.fasta = ""
curchn.sstruc = ""
curchn.start = ii
curchn.chainid = chainid [ chindx ]
curchn.mute = 0
end
curchn.fasta = curchn.fasta .. protNfo.fastac [ ii ]
--
-- secondary structure codes don't need translation,
-- so we don't have the equivalent of "fastac" here,
-- just use the codes from the Foldit API
--
curchn.sstruc = curchn.sstruc .. protNfo.ss [ ii ]
if protNfo.mute [ ii ] then
curchn.mute = curchn.mute + 1
end
if protNfo.last [ ii ] then
curchn.stop = ii
curchn.len = curchn.stop - ( curchn.start - 1 )
end
end
for ii = 1, #chainz do
chainz [ ii ].fastab = chainz [ ii ].fasta
chainz [ ii ].sstrucb = chainz [ ii ].sstruc
end
return chainz
end
function report_time(start_clock,start_time,clock_msg,time_msg)
local seconds,minutes,hours,days
if clock_msg==nil then clock_msg="CPU time" end
if time_msg==nil then time_msg="Elasped time" end
print(string.format("%s",os.date()))
days,remainder=math.modf((os.clock()-start_clock)/(24*60*60))
hours,remainder=math.modf(remainder*24)
minutes,remainder=math.modf(remainder*60)
seconds,remainder=math.modf(remainder*60)
print(string.format("%s(%02id:%02ih:%02im:%02is)",clock_msg,days,hours,minutes,seconds))
days,remainder=math.modf(os.difftime(os.time(),start_time)/(24*60*60))
hours,remainder=math.modf(remainder*24)
minutes,remainder=math.modf(remainder*60)
seconds,remainder=math.modf(remainder*60)
print(string.format("%s(%02id:%02ih:%02im:%02is)",time_msg,days,hours,minutes,seconds))
end
function getSS ( )
ssList = ""
for ii = 1, structure.GetCount () do
ssList = ssList .. structure.GetSecondaryStructure ( ii )
end
return ssList
end
function setSS ( chain )
local changes = 0
local errz = 0
local offset = chain.start - 1
local sstrucn = "" -- possibly changed chain
for ii = 1, chain.stop - ( chain.start - 1 ) do
local sType = chain.sstruc:sub ( ii, ii )
local oType = chain.sstrucb:sub ( ii, ii )
--
-- unlike AAEdit,
-- no conversion of sType is needed here,
-- assuming it will normally be H/E/L,
-- or maybe M for ligand
--
-- also, H/E/L codes can usually be changed,
-- but ligand code M can't
--
-- (nevertheless, we'll let the user have a
-- go at it, see what happens)
--
if sType ~= oType then
local sName = SSNames [ sType ]
if sName ~= nil then
structure.SetSecondaryStructure ( ii + offset, sType )
local newss = structure.GetSecondaryStructure ( ii + offset )
if newss == sType then
changes = changes + 1
sstrucn = sstrucn .. sType
else
print ( "segment "
.. ii + offset ..
" ("
.. chain.chainid .. ":" .. ii ..
") change to type \""
.. sType .. "\" failed" )
errz = errz + 1
sstrucn = sstrucn .. oType
end
else
print ( "segment "
.. ii + offset ..
" ("
.. chain.chainid .. ":" .. ii ..
"), skipping invalid type \""
.. sType ..
"\"" )
errz = errz + 1
sstrucn = sstrucn .. oType
end
else
sstrucn = sstrucn .. oType
end
end
chain.sstruc = sstrucn
chain.sstrucb = sstrucn
return changes, errz
end
function GetParameters ( chnz )
local dlog = dialog.CreateDialog ( ReVersion )
dlog.sc0 = dialog.AddLabel ( "segment count = " .. structure.GetCount () )
local cwd = "chain"
if #chnz > 1 then
cwd = "chains"
end
dlog.chz = dialog.AddLabel ( #chnz .. " " .. cwd )
for ii = 1, #chnz do
local chain = chnz [ ii ]
dlog [ "chn" .. ii .. "l1" ] = dialog.AddLabel (
"Chain "
.. chain.chainid ..
" ("
.. Ctypes [ chnz [ ii ].ctype ] ..
")"
)
dlog [ "chn" .. ii .. "l2" ] = dialog.AddLabel (
"segments "
.. chain.start ..
"-"
.. chain.stop ..
", mutables = "
.. chain.mute ..
", length = "
.. chain.len
)
dlog [ "chn" .. ii .. "ss" ] = dialog.AddTextbox ( "sec struct", chain.sstruc )
end
dlog.u0 = dialog.AddLabel ( "" )
dlog.u1 = dialog.AddLabel ( "Usage: use select all and copy, cut, or paste" )
dlog.u2 = dialog.AddLabel ( "to save or change secondary structure" )
dlog.w0 = dialog.AddLabel ( "" )
dlog.w1 = dialog.AddLabel ( "Windows: ctrl + a = select all" )
dlog.w2 = dialog.AddLabel ( "Windows: ctrl + x = cut" )
dlog.w3 = dialog.AddLabel ( "Windows: ctrl + c = copy" )
dlog.w4 = dialog.AddLabel ( "Windows: ctrl + v = paste" )
dlog.z0 = dialog.AddLabel ( "" )
dlog.ok = dialog.AddButton ( "Change" , 1 )
dlog.exit = dialog.AddButton ( "Exit" , 0 )
if ( dialog.Show ( dlog ) > 0 ) then
for ii = 1, #chnz do
chnz [ ii ].sstruc = ( dlog [ "chn" .. ii .. "ss" ].value:upper ()):sub ( 1, chnz [ ii ].len )
end
return true
else
return false
end
end
function main ()
print ( ReVersion )
print ( "Puzzle: " .. puzzle.GetName () )
print ( "Track: " .. ui.GetTrackName () )
undo.SetUndo ( false )
protNfo.setNfo ()
local changeNum = 0
local ssList = ""
local chnTbl = {} -- chains as table of tables
chnTbl = getChains ()
local cwd = "chain"
if #chnTbl > 1 then
cwd = "chains"
end
print ( #chnTbl .. " " .. cwd )
for ii = 1, #chnTbl do
local chain = chnTbl [ ii ]
if chain.stop == nil then
chain.stop = 999999
end
print ( "chain " .. chain.chainid .. ", start = " .. chain.start .. ", end = " .. chain.stop )
print ( chain.sstruc )
end
while GetParameters ( chnTbl ) do
local cmods = 0
for ii = 1, #chnTbl do
local chain = chnTbl [ ii ]
if chain.sstruc ~= chain.sstrucb then
print ( "--" )
print ( "chain " .. chain.chainid .. " changed" )
cmods = cmods + 1
local old = chain.sstrucb
changeNum = changeNum + 1
local start_time = os.time ()
behavior.SetFiltersDisabled ( true )
local sChg, sErr = setSS ( chnTbl [ ii ] )
behavior.SetFiltersDisabled ( false )
print ( "segments changed = " .. sChg .. ", errors = " .. sErr )
print ( "old chain " .. chain.chainid .. ": " )
print ( old )
print ( "new chain " .. chain.chainid .. ": " )
print ( chain.sstrucb )
end
end
if cmods == 0 then
print ( "--" )
print ( "nothing changed" )
end
end
cleanup ()
end
function cleanup ( errmsg )
--
-- optionally, do not loop if cleanup causes an error
-- (any loop here is automatically terminated after a few iterations, however)
--
if CLEANUPENTRY ~= nil then
return
end
CLEANUPENTRY = true
print ( "---" )
--
-- model 100 - print recipe name, puzzle, track, time, score, and gain
--
local reason
local start, stop, line, msg
if errmsg == nil then
reason = "complete"
else
--
-- model 120 - civilized errmsg reporting,
-- thanks to Bruno K. and Jean-Bob
--
start, stop, line, msg = errmsg:find ( ":(%d+):%s()" )
if msg ~= nil then
errmsg = errmsg:sub ( msg, #errmsg )
end
if errmsg:find ( "Cancelled" ) ~= nil then
reason = "cancelled"
else
reason = "error"
end
end
print ( ReVersion .. " " .. reason )
print ( "Puzzle: " .. puzzle.GetName () )
print ( "Track: " .. ui.GetTrackName () )
if reason == "error" then
print ( "Unexpected error detected" )
print ( "Error line: " .. line )
print ( "Error: \"" .. errmsg .. "\"" )
end
behavior.SetFiltersDisabled ( false )
end
xpcall ( main, cleanup )
