Unsupervised biodiversity estimation using proteomic fingerprints fromMALDI‐TOF MSdata

Abstract

Species identification using matrix assisted laser desorption/ionization time-of-flight mass spectrometry (MALDI-TOF MS) data strongly relies on reference libraries to differentiate species. Because comprehensive reference libraries, especially for metazoans, are rare, we explored the accuracy of unsupervised diversity estimations of communities using MALDI-TOF MS data in the absence of reference libraries to provide a method for future application in ecological research. To discover the best analysis strategy providing high congruence with true community structures, we carried out a simulation with more than 30,000 analyses using different combinations of data transformations, dimensionality reductions, and cluster algorithms. Species profile, Hellinger, and presence/absence transformations were applied to raw data and dimensions were reduced using principal component analysis (PCA), t-distributed stochastic neighbor embedding, and uniform manifold approximation and projection. To estimate biodiversity, data were clustered making use of partitioning around medoids, modelbased clustering, and K-means clustering. The analyses were carried out on published mass spectrometry data of harpacticoid copepods. Most successful combinations (Hellinger transformation + PCA or raw data + partitioning around medoids) returned good values even for difficult species distributions containing numerous singleton species. Nevertheless, errors occurred most frequently because of such singleton taxa. Hence, replicative sampling in wide sampling areas for analysis is emphasized to increase the minimum number of specimens per species, thus reducing putative sources of errors. Our results demonstrate that MALDI-TOF MS data can be used to accurately estimate the biodiversity of unknown communities using unsupervised learning methods. The provided approach allows the biodiversity comparison of sampled regions for which no reference libraries are available. Hence, especially data on groups which demand a time-consuming identification or are highly abundant can be analyzed within short working time, accelerating ecological studies.

Proteomic fingerprinting facilitates biodiversity assessments in understudied ecosystems: A case study on integrated taxonomy of deep sea copepods

“…The intercomparison of MOTUs and POTUs revealed that 7% of specimens were assigned differently. A slightly smaller error rate of 4% has been observed applying clustering on simulated data sets (Rossel and Martínez Arbizu, 2020). Misidentification of single specimens was neither detectable using a direct comparison of sample distances nor by consensus clustering, indicating that misidentification is probably caused by variance in the proteomic profiles.…”

“…Yet, no gold standard for unsupervised species delimitation has been developed. A previous study successfully applied partitioning around medoids (PAM) clustering in combination with the silhouette index to predict species number (Rossel and Martínez Arbizu, 2020). PAM and also k‐means clustering were not applicable to our study due to the expected unbalanced data set with most probably many singletons and small sample size.…”

“…The use of unsupervised classification methods with proteomic fingerprints is a promising technique that may allow a prediction of species diversity without prior species identification (Rossel & Martínez Arbizu, 2020). It implies that specimens can be separated on species level using the dissimilarity of their MALDI spectra as the only source of information.…”

“…However, in the benthopelagic layers of the deep sea, we expect a high number of new, undescribed species, making the establishment of a such a reference library rather time‐ and cost‐intensive. The use of unsupervised classification methods with proteomic fingerprints is a promising technique that may allow a prediction of species diversity without prior species identification (Rossel & Martínez Arbizu, 2020). It implies that specimens can be separated on species level using the dissimilarity of their MALDI spectra as the only source of information.…”

Toward a reliable assessment of potential ecological impacts of deep‐sea polymetallic nodule mining on abyssal infauna

“…Also, DNA sequences and MALDI‐TOF libraries can be simultaneously generated by splitting specimens (if their size allows this) into multiple pieces using some for genomic and some for proteomic fingerprinting (Rossel and Arbizu 2018; Rossel et al 2019; Rossel and Martínez Arbizu 2019). If no libraries for a region are available at all, automatic biodiversity estimation based on a clustering analysis and automatic cluster recognition can be performed (Rossel and Martínez Arbizu 2020) to allow biodiversity comparisons of different investigated areas. A combination of both approaches can be used for areas with an incomplete reference library.…”