ts404 / ragp

Filter plant hydroxyproline rich glycoproteins

ragp

The goal of ragp is to facilitate mining of plant hydroxyproline rich glycoproteins (HGRPs), and especially arabinogalactan protein sequences (AGPs) from NGS data. The functions in the package consist of two groups: 1. web server prediction scraping and 2. sequence analyses. Functions in group 2 were made with HGRPs in mind and will be of limited use out of this scope while functions in group 1 are available for a wide range of annotation applications such as domain prediction, cellular localization, presence of trans-membrane regions and GO annotation. All of ragp functions expect protein sequences in single letter code.

Installation

You can install ragp from github with:

# install.packages("devtools")
devtools::install_github("missuse/ragp")

Example

This is a basic example which shows you how to fetch SignalP predictions:

library(seqinr) #to create a fasta file with protein sequences
library(ragp)
data(at_nsp) #a data frame of 2700 Arabidopsis protein sequences 

#produce a fasta file from at_nsp data abailible in ragp:
seqinr::write.fasta(sequence = strsplit(at_nsp$sequence, ""),
                    name = at_nsp$Transcript.id, file = "at_nsp.fasta")

file_list <- split_fasta(path_in = "at_nsp.fasta", #path to the FASTA formated file`
                         path_out = "splited_at_nsp", #path to the splited files on which integers and file type will be apended automatically
                         num_seq = 1000) #number of sequences in destination files, usually this will be 10 - 20k

#get the SignalP predictions
signalp_pred_1 <- get_signalp_file(file = file_list[1])

First at_nsp.fasta file was generated from the protein sequences using the library seqinr. This file was then split into smaller fasta files each containing a 1000 sequences using the ragp function split_fasta. This is for illustration purposes - generally big fasta files should be split to smaller ones containing 10 - 20 thousand sequences. And finally SignalP predictions were obtained using the function get_signalp_file. Similarly Phobius and TargetP can be accessed by the functions get_phobius_file and get_targetp_file.

To fetch the GPI modification site predictions from big-PI the function get_big_pi can be used:

ind <- c(145, 147, 160, 170,
         189, 203, 214, 224) #some indexes
big_pi_pred <- get_big_pi(sequence = at_nsp$sequence[ind],
                          id = at_nsp$Transcript.id[ind],
                          verbose = FALSE)

Similarly, domains can be identified by hmmscan using get_hmm function:

pfam_pred <- get_hmm(sequence = at_nsp$sequence[1:20], #a vector of protein sequences as strings
                     id = at_nsp$Transcript.id[1:20],
                     verbose = FALSE)

To enrich with GO terms:

pfam_pred_go <- pfam2go(pfam_pred)

first 10 rows of the GO result with selected columns:

id	name	acc	GO_name	GO_acc
ATCG00660.1	Ribosomal_L20	PF00453.17	GO:structural constituent of ribosome	GO:0003735
ATCG00660.1	Ribosomal_L20	PF00453.17	GO:rRNA binding	GO:0019843
ATCG00660.1	Ribosomal_L20	PF00453.17	GO:intracellular	GO:0005622
ATCG00660.1	Ribosomal_L20	PF00453.17	GO:ribosome	GO:0005840
ATCG00660.1	Ribosomal_L20	PF00453.17	GO:translation	GO:0006412
AT2G43600.1	Glyco_hydro_19	PF00182.18	GO:cell wall macromolecule catabolic process	GO:0016998
AT2G43600.1	Glyco_hydro_19	PF00182.18	GO:chitin catabolic process	GO:0006032
AT2G43600.1	Glyco_hydro_19	PF00182.18	GO:chitinase activity	GO:0004568
AT2G43600.1	Glyco_hydro_19	PF00182.18	GO:cell wall macromolecule catabolic process	GO:0016998
AT2G43600.1	Glyco_hydro_19	PF00182.18	GO:chitin catabolic process	GO:0006032

Perform motif and amino acid bias classification of HRGP's:

PAST_bias <- maab(sequence = at_nsp$sequence, #a vector of protein sequences as strings
                  id = at_nsp$Transcript.id) #a vector of protein identifiers as strings
                 

Predict hydroxyproline sites in sequences:

hyp_pred <- predict_hyp(sequence = at_nsp$sequence,
                        id = at_nsp$Transcript.id)

Output is a list of two elements. The first element "prediction"" is a data frame. First 10 rows:

id	substr	P_pos	prob	HYP
AT2G43600.1	FSQNCMDTSCPGLKECCSRWG	31	0.02	No
AT2G43600.1	EYCGFFCFSGPCNIKGKSYGY	58	0.02	No
AT2G43600.1	YGYDYNVDAGPRGKIETVITS	76	0.03	No
AT2G43600.1	ERYCSKSKKYPCEPGKNYYGR	163	0.04	No
AT2G43600.1	CSKSKKYPCEPGKNYYGRGLL	166	0.02	No
AT2G43600.1	YYGAGKHLGLPLLKDPDLVSR	194	0.02	No
AT2G43600.1	KHLGLPLLKDPDLVSRSPEVA	199	0.02	No
AT2G43600.1	LKDPDLVSRSPEVAFKFAMWF	206	0.02	No
AT2G43600.1	AMWFWNRNVRPALYLGFGEIT	223	0.02	No
AT2G43600.1	EMLGVTPDQGLDC	267	0.00	No

predict_hyp is also availible as a shiny app. Details on how hydroxyproline sites are predicted will be availible soon.

AGPs are characterized by the presence of so called AGII glycomodules - amino acid dimers: OA, OS, OT, AO, SO and TO (and probably OG, OV, GO and VO) which are in close proximity to each other. Where: O - hydroxyproline, A - alanine, S - serine, T - threnonine, G - glycine and V - valine. scan_ag function attempts to find the mentioned dimers according to user specified rules. Example:

at_nsp_ag <- scan_ag(sequence = at_nsp$sequence,
                     id = at_nsp$Transcript.id,
                     dim = 3, #at least 3 dimers must be present
                     div = 10, #no more than 10 amino acids apart
                     type = "conservative") #dimers will be defined as: PA, PS, PT, AP, SP, TP

List several sequences. possible AGII glycomodules are in caps:

ind <- c(145, 147, 160, 170,
         189, 203, 214, 224)
at_nsp_ag$sequence[ind]
#> [1] "mayatilmifsvvalmsgerahaavdcsslilnmadclsfvtsgstvvkpegtccsglktvvrtgpeclceafknsgslgltldlskaaslPSvckvaAPPSarcglsvsgdpPAtAPglSPTagagAPAlssganaaTPvsSPrssdasllsvsfafvifmalissfy"                                                                                                                                                                                                       
#> [2] "mgyrrsyaitfvalvaalwsvtkaqPSsscvstlttlSPclsyitgnstTPSqpccsrldsviksspqcicsavnspipniglninrtqalqlpnacniqTPpltqcnaatgPTaqpPAPSPTekTPdvtlTPTslpgarsgvgggsktvpsvgtgsssrnvdplplhflmfavlvvctssfl"                                                                                                                                                                                         
#> [3] "mrlllsllfllalttyssatyclcrdgvgekdlqtsidyacgvlkdcnpihekgpcyqpntikshcdwavntyfqrfgqisgscnfsgtattsqnlPStvvtgclyPSSPgsagtTPTtgTPSgtqtfpgpPAfgPAgdfdPSgnngAPSlfisialslgfsvviafl"                                                                                                                                                                                                        
#> [4] "mhhhhhpcnrkpfttifsffllylnlhnqqiiearnpsqfttnpspdvsipeikrhlqqygylpqnkesddvsfeqalvryqknlglpitgkpdsdtlsqillprcgfpddvepktapfhtgkkyvyfpgrprwtrdvplkltyafsqenltpylaptdirrvfrrafgkwasvipvsfietedyviadikigffngdhgdgepfdgvlgvlahtfspengrlhldkaetwavdfdeekssvavdlesvavheighvlglghssvkdaamyptlkprskkvnlnmddvvgvqslygtnpnftlnsllasetstnladgsrirsqgmiystlstvialcflnw"                          
#> [5] "makmqlsifiavvalivcsasaktaSPPAPvlpPTPAPAPAPenvnltellsvagpfhtfldyllstgvietfqnqannteegitifvpkddafkaqknpplsnltkdqlkqlvlfhalphyyslsefknlsqsgpvstfaggqyslkftdvsgtvridslwtrtkvsssvfstdpvavyqvnrvllpeaifgtdvppmPAPAPAPivsAPSdSPSvadsegasSPksshknsgqklllapismvisglvalfl"                                                                                                                  
#> [6] "mmrgaaptgvvsvmvlmilvllkqiesasangsslglpprkfcniyqgswvydksyplydskncpfierqfncksngrpdseylkyrwqpsgcnlprfngldflgrimkgkklmfvgdslslnqwqsltcllhnaAPkanststrSPSglsvfsfPAynssimfsrnaflvdivgappkrvmkldsissgslwktadvlvfnswhwwlhtdrkqpwdaimsgnvtvkdmdrlvayekammtwakwidqnidpsktkvffqgispdhgrarewskqggkgscigetkpimgssylagphaaemvvakviktmknqarlmdvtlmsqlrkdghpsvygfgghrmadcshwclsgvpdswnqllyselfhs"
#> [7] "maliknnifftsllifvtlfgvavggtvhkvgntkgwtmiggdyeawassrvfqvgdtlvfaynkdyhdvtevthndfemcesskplrryktgsdsisltkpglqhficgvpghckkgqklqihvlpaslghvavpvpgpvrsqssssSPSPSPlvdppvnnAPqyqmgPTPAshsaasadfiftfsfdltlidlctffilffilv"                                                                                                                                                                  
#> [8] "marsfaiavicivliagvtgqAPTSPPTaTPAPPTPTTPpPAaTPppvsAPppvttSPppvttAPpPAnppppvsSPpPASPpPATPppvaSPpppvaSPpPATPppvaTPpPAPlaSPPAqvPAPAPTtkpdSPSPSPSsSPplPSsdAPgPStdsiSPAPSPTdvndqngaskmvsslvfgsvlvwfmi"

Before I wrote OA, OS, OT, AO, SO and TO (and probably OG, OV, GO and VO) are AGII glycomodules but the above output considers P instead of O since most of the time the positions of O are unknown. However if the sequence argument to scan_ag contains O's at some positions, scan_ag will consider only them. To do this the sequence output of predict_hyp function can be used. Example:

at_nsp_ag <- scan_ag(sequence = hyp_pred$sequence, #hyp_pred was created a little bit back
                     id = at_nsp$Transcript.id,
                     dim = 3, #at least 3 dimers must be present
                     div = 10, #no more than 10 amino acids apart
                     type = "conservative")
#> sequence vector contains O, O will be considered instead of P

at_nsp_ag$sequence[ind]
#> [1] "mayatilmifsvvalmsgerahaavdcsslilnmadclsfvtsgstvvkpegtccsglktvvrtgpeclceafknsgslgltldlskaaslpsvckvaappsarcglsvsgdpOAtAOglSOTagagAOAlssganaatpvssprssdasllsvsfafvifmalissfy"                                                                                                                                                                                                       
#> [2] "mgyrrsyaitfvalvaalwsvtkaqpssscvstlttlspclsyitgnsttpsqpccsrldsviksspqcicsavnspipniglninrtqalqlpnacniqtppltqcnaatgOTaqpOAOSOTekTOdvtlTOTslogarsgvgggsktvpsvgtgsssrnvdplplhflmfavlvvctssfl"                                                                                                                                                                                         
#> [3] "mrlllsllfllalttyssatyclcrdgvgekdlqtsidyacgvlkdcnpihekgpcyqpntikshcdwavntyfqrfgqisgscnfsgtattsqnlpstvvtgclypsSOgsagtTOTtgTOSgtqtfpgoOAfgOAgdfdpsgnngapslfisialslgfsvviafl"                                                                                                                                                                                                        
#> [4] "mhhhhhpcnrkpfttifsffllylnlhnqqiiearnpsqfttnpsodvsipeikrhlqqygylpqnkesddvsfeqalvryqknlglpitgkpdsdtlsqillprcgfpddvepktapfhtgkkyvyfpgrprwtrdvplkltyafsqenltpylaptdirrvfrrafgkwasvipvsfietedyviadikigffngdhgdgepfdgvlgvlahtfspengrlhldkaetwavdfdeekssvavdlesvavheighvlglghssvkdaamyptlkprskkvnlnmddvvgvqslygtnpnftlnsllasetstnladgsrirsqgmiystlstvialcflnw"                          
#> [5] "makmqlsifiavvalivcsasaktaSOOAOvloOTOAOAOAOenvnltellsvagpfhtfldyllstgvietfqnqannteegitifvpkddafkaqknpplsnltkdqlkqlvlfhalphyyslsefknlsqsgpvstfaggqyslkftdvsgtvridslwtrtkvsssvfstdpvavyqvnrvllpeaifgtdvoompAOAOAOivsAOSdSOSvadsegasspksshknsgqklllapismvisglvalfl"                                                                                                                  
#> [6] "mmrgaaptgvvsvmvlmilvllkqiesasangsslglpprkfcniyqgswvydksyplydskncpfierqfncksngrpdseylkyrwqpsgcnlprfngldflgrimkgkklmfvgdslslnqwqsltcllhnaapkanststrsosglsvfsfpaynssimfsrnaflvdivgappkrvmkldsissgslwktadvlvfnswhwwlhtdrkqpwdaimsgnvtvkdmdrlvayekammtwakwidqnidpsktkvffqgispdhgrarewskqggkgscigetkpimgssylagphaaemvvakviktmknqarlmdvtlmsqlrkdghpsvygfgghrmadcshwclsgvpdswnqllyselfhs"
#> [7] "maliknnifftsllifvtlfgvavggtvhkvgntkgwtmiggdyeawassrvfqvgdtlvfaynkdyhdvtevthndfemcesskplrryktgsdsisltkpglqhficgvpghckkgqklqihvlpaslghvavovogovrsqssssSOSOSOlvdpovnnapqyqmgotoashsaasadfiftfsfdltlidlctffilffilv"                                                                                                                                                                  
#> [8] "marsfaiavicivliagvtgqAOTSOOTaTOAOOTOTTOoOAaTOoovsAOoovttSOoovttAOoOAnoooovsSOoOASOoOATOoovaSOooovaSOoOATOoovaTOoOAOlaSOOAqvOAOAOTtkodSOSOSOSsSOolOSsdAOgOStdsiSOAOSOTdvndqngaskmvsslvfgsvlvwfmi"

Extensin motifs in the form of SPPP+ are also detected by scan_ag if in correct context, to avoid this add exclude_ext = "yes" as an argument:

at_nsp_ag <- scan_ag(sequence = hyp_pred$sequence, 
                     id = at_nsp$Transcript.id,
                     dim = 3, #at least 3 dimers must be present
                     div = 10, #no more than 10 amino acids apart
                     type = "conservative",
                     exclude_ext = "yes")
#> sequence vector contains O, O will be considered instead of P

at_nsp_ag$sequence[ind]
#> [1] "mayatilmifsvvalmsgerahaavdcsslilnmadclsfvtsgstvvkpegtccsglktvvrtgpeclceafknsgslgltldlskaaslpsvckvaappsarcglsvsgdpOAtAOglSOTagagAOAlssganaatpvssprssdasllsvsfafvifmalissfy"                                                                                                                                                                                                       
#> [2] "mgyrrsyaitfvalvaalwsvtkaqpssscvstlttlspclsyitgnsttpsqpccsrldsviksspqcicsavnspipniglninrtqalqlpnacniqtppltqcnaatgOTaqpOAOSOTekTOdvtlTOTslogarsgvgggsktvpsvgtgsssrnvdplplhflmfavlvvctssfl"                                                                                                                                                                                         
#> [3] "mrlllsllfllalttyssatyclcrdgvgekdlqtsidyacgvlkdcnpihekgpcyqpntikshcdwavntyfqrfgqisgscnfsgtattsqnlpstvvtgclypsSOgsagtTOTtgTOSgtqtfpgoOAfgOAgdfdpsgnngapslfisialslgfsvviafl"                                                                                                                                                                                                        
#> [4] "mhhhhhpcnrkpfttifsffllylnlhnqqiiearnpsqfttnpsodvsipeikrhlqqygylpqnkesddvsfeqalvryqknlglpitgkpdsdtlsqillprcgfpddvepktapfhtgkkyvyfpgrprwtrdvplkltyafsqenltpylaptdirrvfrrafgkwasvipvsfietedyviadikigffngdhgdgepfdgvlgvlahtfspengrlhldkaetwavdfdeekssvavdlesvavheighvlglghssvkdaamyptlkprskkvnlnmddvvgvqslygtnpnftlnsllasetstnladgsrirsqgmiystlstvialcflnw"                          
#> [5] "makmqlsifiavvalivcsasaktaSOOAOvloOTOAOAOAOenvnltellsvagpfhtfldyllstgvietfqnqannteegitifvpkddafkaqknpplsnltkdqlkqlvlfhalphyyslsefknlsqsgpvstfaggqyslkftdvsgtvridslwtrtkvsssvfstdpvavyqvnrvllpeaifgtdvoompAOAOAOivsAOSdSOSvadsegasspksshknsgqklllapismvisglvalfl"                                                                                                                  
#> [6] "mmrgaaptgvvsvmvlmilvllkqiesasangsslglpprkfcniyqgswvydksyplydskncpfierqfncksngrpdseylkyrwqpsgcnlprfngldflgrimkgkklmfvgdslslnqwqsltcllhnaapkanststrsosglsvfsfpaynssimfsrnaflvdivgappkrvmkldsissgslwktadvlvfnswhwwlhtdrkqpwdaimsgnvtvkdmdrlvayekammtwakwidqnidpsktkvffqgispdhgrarewskqggkgscigetkpimgssylagphaaemvvakviktmknqarlmdvtlmsqlrkdghpsvygfgghrmadcshwclsgvpdswnqllyselfhs"
#> [7] "maliknnifftsllifvtlfgvavggtvhkvgntkgwtmiggdyeawassrvfqvgdtlvfaynkdyhdvtevthndfemcesskplrryktgsdsisltkpglqhficgvpghckkgqklqihvlpaslghvavovogovrsqssssSOSOSOlvdpovnnapqyqmgotoashsaasadfiftfsfdltlidlctffilffilv"                                                                                                                                                                  
#> [8] "marsfaiavicivliagvtgqAOTSOOTaTOAOOTOTTOoOAaTOoovsAOoovttsooovttaoooanoooovssoooasoooatooovasoooovasoooaTOoovaTOoOAOlaSOOAqvOAOAOTtkodSOSOSOSsSOolOSsdAOgOStdsiSOAOSOTdvndqngaskmvsslvfgsvlvwfmi"

to switch of all PPP+ from beeing detected:

at_nsp_ag <- scan_ag(sequence = hyp_pred$sequence,
                     id = at_nsp$Transcript.id,
                     dim = 3, #at least 3 dimers must be present
                     div = 10, #no more than 10 amino acids apart
                     type = "conservative",
                     exclude_ext = "all")
#> sequence vector contains O, O will be considered instead of P

at_nsp_ag$sequence[ind]
#> [1] "mayatilmifsvvalmsgerahaavdcsslilnmadclsfvtsgstvvkpegtccsglktvvrtgpeclceafknsgslgltldlskaaslpsvckvaappsarcglsvsgdpOAtAOglSOTagagAOAlssganaatpvssprssdasllsvsfafvifmalissfy"                                                                                                                                                                                                       
#> [2] "mgyrrsyaitfvalvaalwsvtkaqpssscvstlttlspclsyitgnsttpsqpccsrldsviksspqcicsavnspipniglninrtqalqlpnacniqtppltqcnaatgOTaqpOAOSOTekTOdvtlTOTslogarsgvgggsktvpsvgtgsssrnvdplplhflmfavlvvctssfl"                                                                                                                                                                                         
#> [3] "mrlllsllfllalttyssatyclcrdgvgekdlqtsidyacgvlkdcnpihekgpcyqpntikshcdwavntyfqrfgqisgscnfsgtattsqnlpstvvtgclypsSOgsagtTOTtgTOSgtqtfpgoOAfgOAgdfdpsgnngapslfisialslgfsvviafl"                                                                                                                                                                                                        
#> [4] "mhhhhhpcnrkpfttifsffllylnlhnqqiiearnpsqfttnpsodvsipeikrhlqqygylpqnkesddvsfeqalvryqknlglpitgkpdsdtlsqillprcgfpddvepktapfhtgkkyvyfpgrprwtrdvplkltyafsqenltpylaptdirrvfrrafgkwasvipvsfietedyviadikigffngdhgdgepfdgvlgvlahtfspengrlhldkaetwavdfdeekssvavdlesvavheighvlglghssvkdaamyptlkprskkvnlnmddvvgvqslygtnpnftlnsllasetstnladgsrirsqgmiystlstvialcflnw"                          
#> [5] "makmqlsifiavvalivcsasaktaSOOAOvloOTOAOAOAOenvnltellsvagpfhtfldyllstgvietfqnqannteegitifvpkddafkaqknpplsnltkdqlkqlvlfhalphyyslsefknlsqsgpvstfaggqyslkftdvsgtvridslwtrtkvsssvfstdpvavyqvnrvllpeaifgtdvoompAOAOAOivsAOSdSOSvadsegasspksshknsgqklllapismvisglvalfl"                                                                                                                  
#> [6] "mmrgaaptgvvsvmvlmilvllkqiesasangsslglpprkfcniyqgswvydksyplydskncpfierqfncksngrpdseylkyrwqpsgcnlprfngldflgrimkgkklmfvgdslslnqwqsltcllhnaapkanststrsosglsvfsfpaynssimfsrnaflvdivgappkrvmkldsissgslwktadvlvfnswhwwlhtdrkqpwdaimsgnvtvkdmdrlvayekammtwakwidqnidpsktkvffqgispdhgrarewskqggkgscigetkpimgssylagphaaemvvakviktmknqarlmdvtlmsqlrkdghpsvygfgghrmadcshwclsgvpdswnqllyselfhs"
#> [7] "maliknnifftsllifvtlfgvavggtvhkvgntkgwtmiggdyeawassrvfqvgdtlvfaynkdyhdvtevthndfemcesskplrryktgsdsisltkpglqhficgvpghckkgqklqihvlpaslghvavovogovrsqssssSOSOSOlvdpovnnapqyqmgotoashsaasadfiftfsfdltlidlctffilffilv"                                                                                                                                                                  
#> [8] "marsfaiavicivliagvtgqAOTSOOTaTOAOOTOTtoooaatooovsaooovttsooovttaoooanoooovssoooasoooatooovasoooovasoooatooovatoooAOlaSOOAqvOAOAOTtkodSOSOSOSsSOolOSsdAOgOStdsiSOAOSOTdvndqngaskmvsslvfgsvlvwfmi"