Metabolome-Based Classification of Snake Venoms by Bioinformatic Tools

Abstract

Snakebite is considered a neglected tropical disease, and it is one of the most intricate ones. The variability found in snake venom is what makes it immensely complex to study. These variations are present both in the big and the small molecules found in snake venom. This study focused on examining the variability found in the venom's small molecules (i.e., mass range of 100-1000 Da) between two main families of venomous snakes-Elapidae and Viperidae-managing to create a model able to classify unknown samples by means of specific features, which can be extracted from their LC-MS data and output in a comprehensive list. The developed model also allowed further insight into the composition of snake venom by highlighting the most relevant metabolites of each group by clustering similarly composed venoms. The model was created by means of support vector machines and used 20 features, which were merged into 10 principal components. All samples from the first and second validation data subsets were correctly classified. Biological hypotheses relevant to the variation regarding the metabolites that were identified are also given.

Keywords: data analysis; metabolomics; script-controlled peak integration; venom variation.

Figure 1

Order, family, genus, clade and subclade of the individuals that produced the samples.

Chart 1

Graphic depiction of the workflow followed to develop the classifier model. After the development of the coherent list with all the metabolites and their levels in each venom, two classifiers were used to optimize the dimensionality reduction of this list.

Figure 2

Two-dimensional representation of the scores generated by the model for Rep = 20 and n^o PCs = 10 of PCs 1 and 2. Orange colors represent Viperidae venoms; dark blue colors represent Asian Naja venoms; white colors represent African non-spitting cobra venoms; dark cyan colors represent African spitting cobra venoms; and light cyan colors represent Dendroaspis venoms.

Figure 3

Heatmap of the autoscaled matrix for PCA with Rep = 20. Twenty features were considered relevant, and, in the figure, a correlation between the color and the original data can be seen (lighter colors signify more signal, and darker colors signify less signal). The values represented are the ones from the autoscaled matrix. The subclades were also classified for biological relevance based on their family and genus.

Figure 4

Two-dimensional representation of the Elapidae scores generated by the model for Rep = 20 and n^o PCs = 10 of PCs 2 and 3. Subspecies were also indicated. Dark blue colors represent Asian Naja venoms; white colors represent African non-spitting cobra venoms; dark cyan colors represent African spitting cobra venoms; and light cyan colors represent Dendroaspis venoms. Legend: A: Dendroaspis angusticepsis, B: Dendroaspis jamesoni jamesoni, C: Dendroaspis jamesoni kaimosae, D: Dendroaspis polylepsis, E: Dendroaspis viridis, F: Naja mossambica, G: Naja nigricinta, H: Naja nigricollis, I: Naja nubiae, J: Naja haje, K: Naja legionis, L: Naja nivea, M: Naja subfulva, N: Naja atra, O: Naja kaouthia, P: Naja naja, Q: Naja oxiana.

Figure 5

(a) Two-dimensional representation of PCs 1 and 2 holding the scores generated by the totality of the samples when using the PCs obtained for the model set with Rep = 20 and n^o PCs = 10. Red colors represent Viperidae snakes, while blue and white colors represent Elapidae ones. All the genera, species, subspecies and clades are also shown. (b) Three zoom-in views (i.e., 1, 2 and 3) of the densest parts of a) to further clarify the relative position of each of the samples. Black dots represent Bothrops venoms; yellow dots represent Crotalus venoms in the second validation set; orange dots represent Crotalus venoms in the model dataset; dark blue dots represent Asian Naja venoms; white dots represent African non-spitting cobra venoms; dark cyan dots represent African spitting cobra venoms; and light cyan dots represent Dendroaspis venoms. Legend: A: Dendroaspis angusticepsis, B: Dendroaspis jamesoni jamesoni, C: Dendroaspis jamesoni kaimosae, D: Dendroaspis polylepsis, E: Dendroaspis viridis, F: Naja mossambica, G: Naja nigricinta, H: Naja nigricollis, I: Naja nubiae, J: Naja pallida, K: Naja annulifera, L: Naja haje, M: Naja legionis, N: Naja nivea, O: Naja subfulva, P: Naja atra, Q: Naja kaouthia, R: Naja naja, S: Naja Naja siamensis, T: Naja oxiana, 1: Crotalus adamanteus, 2: Crotalus culminatus, 3: Crotalus durissus cumanensis, 4: Crotalus durissus terrificus, 5: Crotalus vegrandis (1), 6: Crotalus vegrandis (2), 7: Crotalus enyo, 8: Crotalus horridus, 9: Crotalus horridus atricaudatus (1), 10: Crotalus horridus atricaudatus (2), 11: Crotalus lepidus kluberri, 12: Crotalus lepidus maculosus, 13: Crotalus morulus, 14: Crotalus molossus molossus, 15: Crotalus molossus nigricans, 16: Crotalus molossus oaxacus, 17: Crotalus phyrrus (1), 18: Crotalus phyrrus (2), 19: Crotalus cerberus, 20: Crotalus ravus ravus, 21: Crotalus lorenzoensis, 22: Crotalus ruber ruber, 23: Crotalus scutellatus salvini, 24: Crotalus scutellatus scutellatus, 25: Crotalus tortuguensis, 26: Crotalus triseratus triseratus, 27: Crotalus viridis viridis, 28: Bothrops asper, 29: Bothrops alternatus, 30: Bothrops atrox.

Figure 6

Heatmap of the autoscaled complete dataset for PCA with Rep = 20. These 20 features are relevant to the model, and, in the figure, a correlation between the color and the autoscaled data is given (lighter colors signify more signal, and darker colors signify less signal). The family, genus and clade were also classified for biological relevance based on their family and genus.