Fig. 5
From: Development of the autonomous lab system to support biotechnology research

Development of an algorithm to derive the optimized medium conditions for the product amount of the glutamic acid-producing cells. (A) Bayesian-optimized models for the glutamic acid production: (i) Model constructed based on the initial dataset. (ii)–(iv) Bayesian-optimized models successively updated on the glutamic acid production data shown in (Fig. 5B). Additional amounts of the four components (CaCl2, MgSO4, CoCl2, and ZnSO4) into the M9 medium are shown on the X- and Y-axes of the heatmaps. Additional amounts of the four components (CaCl2, MgSO4, CoCl2, and ZnSO4) into the M9 medium are shown on the X- and Y-axes of the heatmaps. Color scales are shown as predicted-relative values of the glutamic acid concentrations (vs. the control) when the cells were cultivated under the adjusted medium conditions. Star symbols indicate the proposed combinations of the four components. (B) Glutamic acid production data obtained from the cells cultivated under the adjusted medium conditions proposed by the Bayesian-optimized models shown in (Fig. 5A). These data were normalized by those of the cells cultivated in the control medium (shown as Norm. Glu conc., mean ± SD, n = 3–6). Significant differences compared to the control were shown statistically by Student’s t-test (* P < 0.05, ** P < 0.01, and *** P < 0.005). These data were obtained automatically under the ANL.