Deep data analytics for genetic engineering of diatoms linking genotype to phenotype via machine learning

Abstract

Genome engineering for materials synthesis is a promising avenue for manufacturing materials with unique properties under ambient conditions. Biomineralization in diatoms, unicellular algae that use silica to construct micron-scale cell walls with nanoscale features, is an attractive candidate for functional synthesis of materials for applications including photonics, sensing, filtration, and drug delivery. Therefore, controllably modifying diatom structure through targeted genetic modifications for these applications is a very promising field. In this work, we used gene knockdown in Thalassiosira pseudonana diatoms to create modified strains with changes to structural morphology and linked genotype to phenotype using supervised machine learning. An artificial neural network (NN) was developed to distinguish wild and modified diatoms based on the SEM images of frustules exhibiting phenotypic changes caused by a specific protein (Thaps3_21880), resulting in 94% detection accuracy. Class activation maps visualized physical changes that allowed the NNs to separate diatom strains, subsequently establishing a specific gene that controls pores. A further NN was created to batch process image data, automatically recognize pores, and extract pore-related parameters. Class interrelationship of the extracted paraments was visualized using a multivariate data visualization tool, called CrossVis, and allowed to directly link changes in morphological diatom phenotype of pore size and distribution with changes in the genotype.

Computational analysis of the effects of nitrogen source and sin1 knockout on biosilica morphology in the model diatom Thalassiosira pseudonana

“…Individual knockdowns of each of the genes sap1, sap2 and sap3 resulted in various aberrations in biosilica morphology [15]. Knockdown of the gene Thaps3_21880 caused a reduction of the pore density while increasing the mean area of the same [16]. Knockout of the silicanin-1 gene (sin1) had affected valve morphology and caused substantially reduced strength and stiffness of the cell wall [17].…”

“…This led to the identification of a phenotype switch from a mesh-like to a tree-like ridge pattern in cells that were depleted of silicic acid, the precursor for silica [26]. Artificial neural networks (NNs) were used to identify in SEM images slight structural differences between the valves from T. pseudonana wild type and the Thaps3_21880 gene knockdown mutant [16]. However, although NNs can be used to determine if any morphological differences exist between multiple groups of images, they do not reveal what the specific morphological differences are.…”

Shedding light on biosilica morphogenesis by comparative analysis of the silica-associated proteomes from three diatom species

“…It is conceivable that the biosynthesis of such an elaborate structure will require a larger number of morphogenetic proteins compared to the comparatively simple patterns of branched, cross-linked ribs and cribrum pores. One of the proteins we found to be fultoportulae-localized, TpSSP20, was previously suggested to be involved in cribrum pore formation based on RNAi experiments targeting the TpSSP20 gene (Trofimov et al, 2019). The resulting mutant phenotype exhibited alterations in the area and density of the cribrum pores in the valve, but no effect on the positioning or morphology of the fultoportulae was reported.…”

Novel nanomaterial–organism hybrids with biomedical potential

“…Mineralized virus is characterized by pH sensitivity, which is beneficial for the construction of a robust vaccine with high immunogenicity. Genetic engineering was used to produce mineralized diatoms, bacteria, yeast cells, and other organisms with unique characteristics (Arakaki et al, 2020; Kwon et al, 2015; Lisse et al, 2017; Trofimov et al, 2019). For example, intracellular mineralization of yeast (Matsumoto et al, 2015) may involve iron‐responsive transcriptional activator Aft1 that regulates the content of iron in the cell.…”