Recipe: NetSurfP 1.1

created by LociOiling

Profile

• Name: NetSurfP 1.1
• ID: 102282
• Shared with: Public
• Parent: NetSurfP 1.0
• Children:
  • NetSurfP 1.2
• Created on: January 14, 2017 at 09:01 AM UTC
• Updated on: January 14, 2017 at 09:01 AM UTC
• Description:

  Convert NetSurfP webpage output into secondary structure prediction and copy-and-paste spreadsheet format. V1.1 handles blank lines gracefully and adds a confidence prediction.

Code

--[[
    NetSurfP - convert NetSurfP format


    NetSurfP, www.cbs.dtu.dk/services/NetSurfP/, takes a primary sequence
    and outputs predicted surface accessibility and secondary structure.

    "Surface accessibility" seems to be more or less the inverse of what's
    called "predicted residue burial" in Foldit. It's the chances that a 
    given residue will be on the outside of a protein.

    NetSurfP outputs most of its values as probabilities, and it uses a 
    columnar format. Unfortunately, the formatting used on the output page
    does not lend itself to being copied and pasted into a spreadsheet.

    This recipe converts the columnar format of NetSurfP output to a 
    tab-delimited format which can be copied and pasted into a spreadsheet. 
    
    The recipe also attempts to create a Foldit secondary structure string 
    from the NetSurfP probabilities.

    version 1.0 -- 2017/01/08 -- LociOiling
    * new recipe
    version 1.1 -- 2017/01/14 -- LociOiling
    * ignore blank lines in input
    * add confidence prediction
]]--

--
-- Globals
--
Recipe = "NetSurfP"
Version = "1.1"
ReVersion = Recipe .. " v." .. Version

--
--  end of globals section
--

function NSPReader ( nspentry )

    local linecnt = 0
    local comments = 0
    local unknown = 0

    local louts = ""    -- whole shebang in spreadsheet format
    local ssp = ""      -- ss prediction
    local ssc = ""      -- ss confidence

--
--  manifest constants for column positions,
--  change these if NetSurfP format changes
--
    local PHELIX = 8
    local PSHEET = 9
    local PLOOP = 10
--
--  column header, also needs attention in input format changes
--
    local cHead =
        "\"burial\"\t\"aa\"\t\"seqnam\"\t\"segnum\"\t\"rsa\"\t\"absaccess\"\t\"zFit\"\t\"pHelix\"\t\"pSheet\"\t\"pLoop\"\n"

    for line in nspentry:gmatch ( "(.-)[\n*\r*]" ) do
        if line ~= nil and line:len () > 0 then
            local pHelix = 0
        local pSheet = 0
        local pLoop = 0

            if line:match ( "#" ) then
                comments = comments + 1
            else
                local lout = ""
                local col = 0
                for toke in line:gmatch ( "[%S]+" ) do
                    if lout:len () > 0 then
                        lout = lout .. "\t"
                    end
                    if tonumber ( toke ) == nil then
                        lout = lout .. "\"" .. toke .. "\""
                    else
                        lout = lout .. toke
                    end
                    col = col + 1

                    if     col == PHELIX then
                        pHelix = toke
                    elseif col == PSHEET then
                        pSheet = toke
                    elseif col == PLOOP then
                        pLoop = toke
                    end
                end
                if col > 0 then
                    lout = lout .. "\n"
                    if louts:len () == 0 then
                        louts = louts .. cHead
                    end
                    louts = louts .. lout
                --
                --  pick highest probability for secondary structure
                --
                    local pred = "L"
                    local prob = pLoop
                    if     pHelix > pLoop then
                        if pHelix > pSheet then
                            pred = "H"
                            prob = pHelix
                        else
                            pred = "E"
                            prob = pSheet
                        end
                    else
                        if pSheet > pLoop then
                            pred = "E"
                            prob = pSheet
                        end
                    end
                    ssp = ssp .. pred
                    if tonumber ( prob ) < 1.0 then
                        ssc = ssc .. string.sub ( tostring ( prob * 10 ), 1, 1 )
                    else
                        ssc = ssc .. "9" -- should never occur
                    end
                end
            end
        end
        linecnt = linecnt + 1
    end

    print ( "number of lines = " .. linecnt )
    print ( "number of comments = " .. comments )

    return louts, ssp, ssc
end

function GetNetSurfP ()
    local dlog = dialog.CreateDialog ( ReVersion )

    local tab = ""
    dlog.tab0 = dialog.AddLabel ( "NetSurfP Output" )
    dlog.tab = dialog.AddTextbox ( "output", tab )
    dlog.u0 = dialog.AddLabel ( "" )
    dlog.u1 = dialog.AddLabel ( "Usage: copy the NetSurfP output for a sequence" ) 
    dlog.u2 = dialog.AddLabel ( "and paste into the output box" )
    dlog.w0 = dialog.AddLabel ( "" )
    
    dlog.ok = dialog.AddButton ( "OK" , 1 )
    dlog.exit = dialog.AddButton ( "Exit" , 0 )

    if ( dialog.Show ( dlog ) > 0 ) then
        tab = dlog.tab.value
        return tab
    else
        return ""
    end
    return tab
end

function ShowResults ( csv, ssp, ssc )
    local dlog = dialog.CreateDialog ( ReVersion )

    dlog.tab0 = dialog.AddLabel ( "NetSurfP Reformatted Output" )
    dlog.csv = dialog.AddTextbox ( "csv", csv )
    dlog.ssp = dialog.AddTextbox ( "SS pred", ssp )
    dlog.ssc = dialog.AddTextbox ( "SS conf", ssc )
    dlog.u0 = dialog.AddLabel ( "" )
    dlog.u1 = dialog.AddLabel ( "csv is \"comma separated values\" for spreadsheet" )
    dlog.u2 = dialog.AddLabel ( "SS pred is secondary structure prediction" )
    dlog.u3 = dialog.AddLabel ( "SS conf is prediction confidence, 1 low, 9 high" ) 
    dlog.w0 = 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 ( "OK" , 1 )

    dialog.Show ( dlog ) 
end

function main ()
    print ( ReVersion )
    print ( "Puzzle: " .. puzzle.GetName () )
    print ( "Track: " .. ui.GetTrackName () )
   
    local   nsp = ""

    nsp = GetNetSurfP ()
    if nsp:len () > 0 then
        local csv = ""
        local ssp = "" 
        local ssc = "" 

        csv, ssp, ssc = NSPReader ( nsp )

        if csv ~= nil and csv:len () > 0 
        and ssp ~= nil and ssp:len () > 0 then
            ShowResults ( csv, ssp, ssc )
            print ( "---spreadsheet format---" )
            print ( csv )
            print ( "---secondary structure prediction---" )
            print ( ssp )
            print ( "---prediction confidence---" )
            print ( ssc )
        else
            print ( "no results, input format may be wrong" )
        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
end

xpcall ( main, cleanup )

Comments

[LociOiling]

NetSurfP 1.1 reformats the secondary structure prediction from NetSurfP to make it more useful to Foldit players.

NetSurfP 1.1 adds secondary prediction confidence to the output of NetSurfP 1.0. For each segment in the input, the confidence ranges from 0 to 9, with 0 being low confidence. The confidence is simply the first digit of the probability of the winning structure prediction for that segment, so 0.994 gives confidence "9", and 0.590 gives "5".

Sample scriptlog output:

---secondary structure prediction---
LHHHHHHHHHHHHHHLLLLLLLHHHHHHHHHHHHHHLLLLLEEEELLLLEEEEEELLLLLLLLLL
---prediction confidence---
96899999999988758888558899999999987668996688758865678856798666789

NetSurfP 1.1 also ignores blank lines in the input. These blank lines may occur when copying from the NetSurfP output page.

See NetSurfP 1.0 for usage.