README.md

In vanbibn/mineDisProt: Mine/Extract data on protein disorder from DisProt (and other databases)

mineDisProt

The package mineDisProt was developed to extract data from unstructured or semi-structured formats and compile into matrices or data frames more amenable to analysis of intrinsically disorder (ID).

In proteins, intrinsic disorder (ID) is a phenomenon that describes the lack of a stable, or ordered, tertiary structure while still maintaining physiologic functions. Informatics tools, such as those provided by PONDR](http://www.pondr.com/) and PONDR-FIT make it easy to analyze a protein sequence for intrinsic disorder; however, this task can become tedious if one has tens or even hundreds of sequences to analyze.

My goal in developing mineDisProt was to simplify the data collection process so you can focus on the analysis of your data set.

Installation

You can install the development version from GitHub with:

# install.packages("devtools")
devtools::install_github("vanbibn/mineDisProt")

Examples

First, we need to load the package.

library(mineDisProt)

extract_pondr()

This function extracts numerical data from text porduced by the VLXT, VL3, and VSL2 disorder predictors from the Predictor of Natural Disordered Regions PONDR. For each protien sequence, the data from the Raw Output of all three predictors were pasted into a .txt file.

pondr_data <- extract_pondr("inst/extdata/pondr_text/")

# here we will view the data from the VLXT predictor
pondr_data[,1:6]

#>        resid.VLXT dis_rgns.VLXT n_dis.VLXT lg_rgn.VLXT pct.VLXT avg.VLXT
#> A4D2B8        440            10        190          49    43.18   0.4250
#> A8MQ11        134             4         48          17    35.82   0.3672
#> B4DYI2       1134            20        704         118    62.08   0.5658
#> H3BSY2        632            12        483         158    76.42   0.6594
#> O14715       1765            34        582          59    32.97   0.3516
#> P0C2Y1        421             8        209          71    49.64   0.4933
#> P0DPF3       1111            17        701         118    63.10   0.5753
#> Q5VU36       1347            28        710         107    52.71   0.5038
#> Q5VYP0       1347            29        714         107    53.01   0.5085
#> Q86XG9        351             5        258          93    73.50   0.6479

* You may get warnings about “incomplete final line found [in the text file].”

extract_pondr.noVL3()

This is a version of extract_pondr(), except the raw data for the VL3 predictor is missing from the raw data text files.

extract_pondr.noVL3("inst/extdata/pondr_text_withoutVL3/")

#>            resid.VLXT dis_rgns.VLXT n_dis.VLXT lg_rgn.VLXT pct.VLXT avg.VLXT
#> A0A087WVF3        549             9        249         104    45.36   0.4753
#> A6NDS4            549             9        247         104    44.99   0.4739
#> D6RF30            607            11        451         206    74.30   0.6678
#> F8WBI6            632             9        452         104    71.52   0.6309
#> H3BPF8            625            10        462         132    73.92   0.6618
#> I6L899            631            10        428         104    67.83   0.6078
#> O60309           1634            18        980         238    59.98   0.5547
#> P0CJ92            632             9        464         157    73.42   0.6521
#> Q6DHY5            549             9        248         104    45.17   0.4738
#> Q96QE4            947            11        483         169    51.00   0.4719
#>            resid.VSL2 dis_rgns.VSL2 n_dis.VSL2 lg_rgn.VSL2 pct.VSL2 avg.VSL2
#> A0A087WVF3        549             8        276         104    50.27   0.5166
#> A6NDS4            549             8        276         104    50.27   0.5197
#> D6RF30            607             4        571         295    94.07   0.8188
#> F8WBI6            632             5        585         305    92.56   0.8027
#> H3BPF8            625             3        586         468    93.76   0.8106
#> I6L899            631             6        574         305    90.97   0.7882
#> O60309           1634            14       1231         792    75.34   0.6865
#> P0CJ92            632             5        576         308    91.14   0.8186
#> Q6DHY5            549             8        276         104    50.27   0.5188
#> Q96QE4            947            10        579         330    61.14   0.6086

extract_pondrFIT()

This function extracts the relevant data from the temporaty URL produced by analyzing a protein sequence in the PONDR-FIT protein disorder meta-predictor. It then calculates average and percent disorder scores from the per-residue scores. Before using this function, URLs should be collected and put in a .csv file with the UniProt ID in the first column and URL in the second.

extract_pondrFIT("inst/extdata/pondrfit-url.csv")

#> # A tibble: 10 x 5
#>    UniprotID url                             meanDisorder percentDisorder length
#>    <chr>     <chr>                                  <dbl>           <dbl>  <int>
#>  1 Q13401    http://original.disprot.org/te~        0.318          0.185     168
#>  2 A8MQ11    http://original.disprot.org/te~        0.346          0.201     134
#>  3 Q6ZUB1    http://original.disprot.org/te~        0.584          0.618    1445
#>  4 P0DKV0    http://original.disprot.org/te~        0.621          0.690    1188
#>  5 Q5VYP0    http://original.disprot.org/te~        0.508          0.516    1347
#>  6 Q5VVP1    http://original.disprot.org/te~        0.515          0.519    1343
#>  7 P0C874    http://original.disprot.org/te~        0.449          0.438     917
#>  8 Q9BSJ1    http://original.disprot.org/te~        0.211          0.0841    452
#>  9 A7E2F4    http://original.disprot.org/te~        0.725          0.861     631
#> 10 H3BSY2    http://original.disprot.org/te~        0.720          0.840     632

* You may get messages about “Parsed with column specification:”.

extract_string()

Extract data on the quantity of protein interactions given by STRING with the minimum number of interactions at medium (0.4), high (0.7), and highest (0.9) confidence.

string_data <- extract_string("inst/extdata/string/")

# here are the interaction data for each protein at the 0.4 confidence level
string_data[,1:6]

#>            num_nodes.0.4 num_edges.0.4 avg_node_degree.0.4
#> A0A087WVF3            25            39                3.12
#> A6NKT7                59           440               14.90
#> A6NMS7                21            44                4.19
#> O14715                59           343               11.60
#> P0DJD0                42           290               13.80
#> P0DJD1                81          1373               33.90
#> Q3BBV0                29            74                5.10
#> Q6ZQQ2                14            26                3.71
#> Q7Z3J3                34           235               13.80
#> Q99666                63           355               11.30
#>            avg_local_clustering_coef.0.4 expected_num_edges.0.4
#> A0A087WVF3                         0.937                     26
#> A6NKT7                             0.819                     83
#> A6NMS7                             0.869                     21
#> O14715                             0.860                     88
#> P0DJD0                             0.839                     60
#> P0DJD1                             0.772                    153
#> Q3BBV0                             0.955                     29
#> Q6ZQQ2                             0.862                     14
#> Q7Z3J3                             0.871                     49
#> Q99666                             0.843                     88
#>            PPI_enrichment_p-value.0.4
#> A0A087WVF3                   9.06e-03
#> A6NKT7                             NA
#> A6NMS7                       9.10e-06
#> O14715                             NA
#> P0DJD0                             NA
#> P0DJD1                             NA
#> Q3BBV0                       1.74e-12
#> Q6ZQQ2                       1.66e-03
#> Q7Z3J3                             NA
#> Q99666                             NA

* You may get some warnings like “NAs introduced by coercion” if your protein is missing data at higher confidence levels.

Note: In future versions of this package, I hope to fully automate the data collection process.

vanbibn/mineDisProt documentation built on April 30, 2020, 1:39 p.m.