MALDIpqi calculates Parchment Glutamine Index from MALDI TOF ZooMS data. It is a sample level measure of the glutamine deamidation from a list of peptides.
For details in the data processing and mathematical method and models to estimate PQI, refer to our publication Nair et al. (2022)
The method was initially developed to estimate parchment quality based on deamidation, following the ideas from Wilson et al. (2012). However, it can be applied in other tissues as in van Doorn et al. (2012) or Brown et al. (2021)
You can install the released version of MALDIpqi from github with:
install.packages('devtools')
devtools::install_github("ismaRP/MALDIpqi")
data_folder = "data/mzML"
results_folder = "results_test"
This is minimal workflow. It assumes the spectra are in data/mzML, in mzML format and with file names in the form “samplename_replicate.ext”. Where the replicate number is 1, 2 or 3 and ext is the extension of the files.
library(MALDIpqi)
params = list(SNR=1.5, iterations=150, hws_smooth=8, halfWindowSize=20)
iso_peaks = getIsoPeaks(
indir=data_folder, outdir=NULL, readf="mzml",
peptides_user = NULL, nchunks = 50, ncores = 1, iocores = 1,
n_isopeaks = 5, min_isopeaks = 4, iterations=150,
smooth_method = "SavitzkyGolay", hws_smooth = 8, halfWindowSize = 20, SNR = 1.5)
q2e = wls_q2e(peptides_user = NULL, n_isopeaks = 5,
data_list = iso_peaks, outdir=results_folder)
# EStimate free gamma model or fixed to -1/2
res_free_gamma = lme_pqi(q2e, outdir=results_folder,
logq = T, n_isopeaks=5, g="free")
res_fixed_gamma = lme_pqi(q2e, outdir=results_folder,
logq = T, n_isopeaks=5, g=-1/2)
If you want to run the samples using the inferred parameters from the
Orval Dataset. You can skip the preprocessing of this dataset, which is
the most time consuming and get the pre-computed isotopic peaks using
get_ref_isopeaks()
. This has been done for different combinations of
parameters, that you can see with get_ref_isopeaks(which_params)
Then get the q2e and the linear mixed effect model
# Extract reference iso_peaks calculated with parms
iso_peaks_orval = get_ref_isopeaks(params=params)
# Calculate q2e
q2e_orval = wls_q2e(peptides_user = NULL, n_isopeaks = 5,
data_list = iso_peaks_orval)
# Estimate LME model
pqi_orval = lme_pqi(q2e_orval, logq = T, g = -1/2, return_model = T)
# Get q2e from new data and log transform
q2e_new = q2e %>% mutate(resp=log(q))
pqi_ucc = predict_pqi(
q2e_new, estimates = pqi_orval$estimates, model = pqi_orval$model,
logq=T) # Here logq=T indicates that new data and model are already log transformed
Nair, B. et al. (2022) ‘Parchment Glutamine Index (PQI): A novel method to estimate glutamine deamidation levels in parchment collagen obtained from low-quality MALDI-TOF data’, bioRxiv. doi:10.1101/2022.03.13.483627.
Wilson, J., van Doorn, N.L. and Collins, M.J. (2012) ‘Assessing the extent of bone degradation using glutamine deamidation in collagen’, Analytical chemistry, 84(21), pp. 9041–9048. https://doi.org/10.1021/ac301333t
van Doorn, N.L. et al. (2012) ‘Site-specific deamidation of glutamine: a new marker of bone collagen deterioration’, Rapid communications in mass spectrometry: RCM, 26(19), pp. 2319–2327. https://doi.org/10.1002/rcm.6351
Brown, S. et al. (2021) ‘Examining collagen preservation through glutamine deamidation at Denisova Cave’, Journal of archaeological science, 133, p. 105454. http://doi.org/10.1016/j.jas.2021.105454
Bethencourt, J.H. et al. (2022) ‘Data from “A biocodicological analysis of the medieval library and archive from Orval abbey, Belgium”’, Journal of open archaeology data, 10(0). Available at: https://doi.org/10.5334/joad.89.
Ruffini-Ronzani, N. et al. (2021) ‘A biocodicological analysis of the medieval library and archive from Orval Abbey, Belgium’, Royal Society Open Science, 8(6), p. 210210. https://doi.org/10.1098/rsos.210210