GitHub

Getting started

Installing NMRTools

The Distributions package is available through the Julia package system by running Pkg.add("NMRTools"). Throughout, we assume that you have installed the package.

The examples in this tutorial also using the Plots package, which can be obtained similarly.

Plot a 1D spectrum

Let's load some example data. This can be a Bruker experiment directory, a specific pdata folder, or an NMRPipe-format file.

using NMRTools, Plots
spec = exampledata("1D_19F")
5000-element NMRData{Float64,1} with dimensions: 
  F1Dim Sampled{Float64} LinRange{Float64}(-119.592, -124.163, 5000) ReverseOrdered Regular Points
 -119.592  4671.52
 -119.593  4675.19
 -119.594  4680.0
 -119.595  4685.86
 -119.596  4692.51
 -119.597  4699.53
 -119.598  4706.36
    ⋮      
 -124.157  5891.3
 -124.158  5885.71
 -124.159  5879.72
 -124.16   5873.64
 -124.161  5867.74
 -124.162  5862.25
 -124.163  5857.34

NMRTools contains Plots recipes for common types of spectrum. To plot our 1D spectrum, we just use the plot command:

plot(spec)

We could zoom in on a particular region using the usual xlims arguments from Plots, but we can also select a chemical shift range from the data directly. To do this, we use square brackets [...] to access the data like an array, but use the .. selector to specify our chemical shift range:

plot(spec[-124.5 .. -123])

All plots can be saved as high quality vector graphics or png files, using the savefig command:

savefig("myspectrum.pdf")

Plot a 2D spectrum

Two-dimensional spectra can be plotted in exactly the same way as for 1Ds.

spec = exampledata("2D_HN")
plot(spec)

Contour levels are set to five times the noise level. The most convenient way to adjust this is simply to multiply or divide the spectrum by some scaling factor. You can also adjust the title - by default taken from the spectrum label - using the title keyword. Use an empty string (title="") to remove the title.

plot(spec / 2, title="spectrum divided by two")

Accessing your data

Spectrum data and associated axis information, metadata, etc, is encapsulated in an NMRData structure.


julia> spec = exampledata("1D_19F")5000-element NMRData{Float64,1} with dimensions: 
  F1Dim Sampled{Float64} LinRange{Float64}(-119.592, -124.163, 5000) ReverseOrdered Regular Points
 -119.592  4671.52
 -119.593  4675.19
 -119.594  4680.0
 -119.595  4685.86
 -119.596  4692.51
 -119.597  4699.53
 -119.598  4706.36
    ⋮
 -124.157  5891.3
 -124.158  5885.71
 -124.159  5879.72
 -124.16   5873.64
 -124.161  5867.74
 -124.162  5862.25
 -124.163  5857.34

Data can be accessed with conventional array indexing, but also using the value-based selectors, Near and ..:

julia> spec[100:105]6-element NMRData{Float64,1} with dimensions: 
  F1Dim Sampled{Float64} LinRange{Float64}(-119.683, -119.687, 6) ReverseOrdered Regular Points
 -119.683  4920.73
 -119.684  4924.97
 -119.685  4928.51
 -119.685  4931.35
 -119.686  4933.45
 -119.687  4934.75
julia> spec[Near(-124)]5808.748229980469
julia> spec[-124 .. -123.5]547-element NMRData{Float64,1} with dimensions: 
  F1Dim Sampled{Float64} LinRange{Float64}(-123.501, -124.0, 547) ReverseOrdered Regular Points
 -123.501  6001.3
 -123.502  6008.11
 -123.502  6015.39
 -123.503  6022.93
 -123.504  6030.49
 -123.505  6037.79
 -123.506  6044.57
    ⋮
 -123.994  5771.73
 -123.995  5773.22
 -123.996  5776.76
 -123.997  5782.32
 -123.998  5789.73
 -123.999  5798.68
 -124.0    5808.75

This also works for multidimensional data. For example:

julia> spec2d = exampledata("2D_HN")1216×512 NMRData{Float64,2} with dimensions: 
  F1Dim Sampled{Float64} LinRange{Float64}(11.0712, 5.72249, 1216) ReverseOrdered Regular Points,
  F2Dim Sampled{Float64} LinRange{Float64}(129.0, 107.043, 512) ReverseOrdered Regular Points
            129.0      128.957   …   107.129    107.086    107.043
 11.0712     20.8555    88.9229      139.835     93.8359    40.3672
 11.0668    -51.7852    12.9678      164.152    139.854    102.899
 11.0624    -94.7793   -53.8916      125.39     129.348    119.876
 11.058    -107.308    -94.2188       62.5664    90.1367   105.768
 11.0536   -117.223   -125.776   …    21.3799    63.6924    96.2686
  ⋮                              ⋱                ⋮
  5.7445    147.816    316.704       288.039    163.503     15.2725
  5.7401    319.171    397.131       -65.3184  -151.356   -236.952
  5.7357    401.177    407.521      -305.667   -354.497   -385.464
  5.73129   348.761    317.303   …  -348.839   -373.441   -371.131
  5.72689   208.9      188.013      -203.875   -225.342   -224.783
  5.72249    55.4658    68.0029      -13.4932   -33.4531   -45.999
julia> spec2d[8.1 .. 8.3, Near(124)]45-element NMRData{Float64,1} with dimensions: 
  F1Dim Sampled{Float64} LinRange{Float64}(8.29779, 8.10409, 45) ReverseOrdered Regular Points
and reference dimensions: 
  F2Dim Sampled{Float64} LinRange{Float64}(124.015, 124.015, 1) ReverseOrdered Regular Points
 8.29779  2439.44
 8.29339  2407.37
 8.28898  2961.7
 8.28458  3801.93
 8.28018  4461.66
 ⋮
 8.1173   1841.04
 8.1129   1772.44
 8.10849  1679.61
 8.10409  1565.78

A plain array of data for the spectrum can be obtained from this using the data command:

julia> data(spec)5000-element Vector{Float64}:
 4671.5184326171875
 4675.192687988281
 4679.9979248046875
 4685.8551025390625
 4692.506896972656
 4699.526428222656
 4706.362731933594
 4712.417419433594
 4717.134460449219
 4720.0810546875
    ⋮
 5900.0340576171875
 5896.165588378906
 5891.2984619140625
 5885.714294433594
 5879.7239990234375
 5873.638427734375
 5867.738952636719
 5862.253173828125
 5857.3370361328125

Similarly, a plain vector containing axis values can be obtained from this using the data command, passing an additional argument to specify the dimension. This can either be a number or the axis type, e.g. F1Dim:

julia> data(spec, 1)5000-element LinRange{Float64, Int64}:
 -119.592, -119.593, -119.594, -119.595, …, -124.161, -124.162, -124.163