Code
--[[
AA Edit - get and set primary structure
The primary structure of a protein is the sequence of the
amino acids that make up the protein.
AA Edit displays the current primary structure as a sequence
of single-character amino acid codes. (Similar codes are
used for DNA and RNA, see "special cases" below.)
The displayed value can be selected and cut or copied.
The string of single-character codes is similar to the
FASTA format accepted by many protein related tools.
FASTA also allows for header information, which some tools,
such as BLAST, require.
If there are any "mutable" segments, the "Change" button
is displayed, and a a new value can be pasted in. If there are
no mutable segments, any input to the "seq" box is ignored.
When the "Change" button is clicked, the currently displayed
primary structure is applied to the protein. The input amino
acid codes are converted to lower case.
The recipe checks each amino acid code against the list of 20
amino acids used in Foldit. Any input not found in the list is
ignored, and the corresponding segment is left unchanged.
Some puzzles have a mix of mutable and non-mutable segments.
The recipe does not attempt to change any non-mutable segments.
If the structure list is longer than the protein, AA Edit
discards the extra entries at the end of the list.
If the structure list is shorter than the protein, AA 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.
special cases
-------------
Some puzzles contain two or more separate protein chains.
The "insulin mutant" puzzle, which appears periodically as a
revisiting puzzle, is an example.
AA Edit detects the beginning and end of a protein chain by
checking the atom count. Each chain is presented separately,
identified by a chain id: "A", "B", "C", and so on.
All the normal rules apply to each chain.
Some puzzles have one or more ligands, each represented by a segment
which returns "x" or "unk" for its amino acid type. This code and
anything else not found in the normal list of 20 amino acids
is changed to "x" for the purposes of this recipe.
Segments with an "x" for their amino acid code in the replacment
string are not changed.
Each ligand is presented as a separate chain.
Very rarely, Foldit puzzles may contain RNA or DNA. These are chains
of nucleobases instead of amino acids. Each segment is one nucleobase.
Foldit uses two-character codes for RNA and DNA. AA Edit translates
these codes into single-character codes. The single-character codes are
ambiguous, for example, RNA adenine is code "ra" in Foldit, and DNA
adenine is "da". Both become "a" externally, which is also used
for alanine in a protein.
AA Edit treats each DNA or RNA section as a separate chain. This allows
it to keep the ambiguous codes straight.
The handling of RNA and DNA has only been tested for RNA. So far, RNA has
only appeared on one science puzzle, and the RNA was not mutable in
that puzzle. DNA has appeared only in intro puzzles, which don't allow
recipes. It's possible that problems may appear if there are ever
for-credit DNA or RNA puzzles again.
An even rarer case was in found in puzzle 879, segment 134, and
puzzle 1378b, segment 30, where an amino acid was modified by
a glycan.
The code "unk" was used for these modified amino acids, but they
did not have the secondary structure code "M" used for ligands.
A modified amino acid like this is treated as protein, and does not break
the amino acid chain.
See "AA Copy Paste Compare v 1.1.1 -- Brow42" for
a full-function recipe that works with primary and
primary structures.
version 1.2 -- 2016/12/23 -- LociOiling
* clone of PS Edit v1.2
* enable 1-step undo with undo.SetUndo ( false )
version 2.0 -- 2018/09/02 -- LociOiling
* detect and report multiple chains
* force filters on at beginning and end
* handle DNA and RNA, use single-letter codes externally
* refine scriptlog output, eliminate timing calls
version 2.0.1 -- 2020/04/16 -- LociOiling
* handle proline at N-terminal correctly
]]--
--
-- Globals
--
Recipe = "AA Edit"
Version = "2.0.1"
ReVersion = Recipe .. " " .. Version
mutable = false -- true if any mutable segments found
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", },
}
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
-- 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.start = ii
curchn.chainid = chainid [ chindx ]
curchn.mute = 0
end
curchn.fasta = curchn.fasta .. protNfo.fastac [ 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
end
return chainz
end
function setChain ( chain )
local changes = 0
local errz = 0
local offset = chain.start - 1
local fastan = "" -- possibly changed chain
for ii = 1, chain.stop - ( chain.start - 1 ) do
local sType = chain.fasta:sub ( ii, ii )
local oType = chain.fastab:sub ( ii, ii )
--
-- for DNA and RNA, convert FASTA to Foldit
--
if chain.ctype == protNfo.DNA then
sType = DNAin [ sType ]
if sType == nil then
sType = protNfo.UNKNOWN_BASE
end
oType = DNAin [ oType ]
if oType == nil then
oType = protNfo.UNKNOWN_BASE
end
elseif chain.ctype == protNfo.RNA then
sType = RNAin [ sType ]
if sType == nil then
sType = protNfo.UNKNOWN_BASE
end
oType = RNAin [ oType ]
if oType == nil then
oType = protNfo.UNKNOWN_BASE
end
end
if sType ~= oType then
local sName = AANames [ sType ]
if sName ~= nil then
if protNfo.mute [ ii + offset ] then
structure.SetAminoAcid ( ii + offset, sType )
local newaa = structure.GetAminoAcid ( ii + offset )
if newaa == sType then
changes = changes + 1
fastan = fastan .. AANames [ sType ] [ AACODE ]
else
print ( "segment "
.. ii + offset ..
" ("
.. chain.chainid .. ":" .. ii ..
") mutation to type \""
.. sType .. "\" failed" )
errz = errz + 1
fastan = fastan .. AANames [ oType ] [ AACODE ]
end
else
print ( "segment "
.. ii + offset ..
" ("
.. chain.chainid .. ":" .. ii ..
") is not mutable, skipping change to type \""
.. sType .. "\"" )
errz = errz + 1
fastan = fastan .. AANames [ oType ] [ AACODE ]
end
else
print ( "segment "
.. ii + offset ..
" ("
.. chain.chainid .. ":" .. ii ..
"), skipping invalid type \""
.. sType ..
"\"" )
errz = errz + 1
fastan = fastan .. AANames [ oType ] [ AACODE ]
end
else
fastan = fastan .. AANames [ oType ] [ AACODE ]
end
end
chain.fasta = fastan
chain.fastab = fastan
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 .. " chains" )
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 .. "ps" ] = dialog.AddTextbox ( "seq", chain.fasta )
end
dlog.u0 = dialog.AddLabel ( "" )
if mutable then
dlog.u1 = dialog.AddLabel ( "Usage: click in text box, " )
dlog.u2 = dialog.AddLabel ( "then use select all and copy, cut, or paste" )
dlog.u3 = dialog.AddLabel ( "to save or change primary structure" )
else
dlog.u1 = dialog.AddLabel ( "Usage: click in text box," )
dlog.u2 = dialog.AddLabel ( "then use select all and copy" )
dlog.u3 = dialog.AddLabel ( "to save primary structure" )
end
dlog.w0 = dialog.AddLabel ( "" )
if mutable then
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" )
else
dlog.w1 = dialog.AddLabel ( "Windows: ctrl + a = select all" )
dlog.w3 = dialog.AddLabel ( "Windows: ctrl + c = copy" )
end
dlog.z0 = dialog.AddLabel ( "" )
if mutable then
dlog.ok = dialog.AddButton ( "Change" , 1 )
end
dlog.exit = dialog.AddButton ( "Exit" , 0 )
if ( dialog.Show ( dlog ) > 0 ) then
for ii = 1, #chnz do
chnz [ ii ].fasta = ( dlog [ "chn" .. ii .. "ps" ].value:lower ()):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 ()
for ii = 1, structure.GetCount () do
if protNfo.mute [ ii ] == true then
mutable = true
break
end
end
local changeNum = 0
local chnTbl = {} -- chains as table of tables
chnTbl = getChains ()
print ( #chnTbl .. " chains" )
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.fasta )
end
while GetParameters ( chnTbl ) do
for ii = 1, #chnTbl do
local chain = chnTbl [ ii ]
if chain.fasta ~= chain.fastab then
print ( "--" )
print ( "chain " .. chain.chainid .. " changed" )
local old = chain.fastab
changeNum = changeNum + 1
local start_time = os.time ()
behavior.SetFiltersDisabled ( true )
local sChg, sErr = setChain ( 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.fastab )
end
end
end
cleanup ()
end
function cleanup ( errmsg )
--
-- do not loop if cleanup causes an error
--
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 )
