Multivariate Data Analysis for Advancing the Interpretation of Bioprocess Measurement and Monitoring Data - Measurement, Monitoring, Modelling, and Control of Bioprocesses - page 180

0.09

AJA018_2.5hours AJA018_3hours

0.08

0.07

0.06

0.05

0.04

0.03

0.02

0.01

0

1 10 19 28 37 46 55 64 73 82 91 100 109 118 127 136 145

Mass (m/z)

Fig. 7 Comparison of PyMS data from the same cultivation using two consecutive time-point
samples in blue and red colour

Fig. 8 Clustering of data from a recombinant E. col cultivation. Triplicates of individual time-
point samples are represented by full circles and connected by lines. Each of the triplicates is
additionally labelled to indicate the fermentation and time post-inoculation

Each PyMS data file consists of 150 mass to ion charge (m/z) data reflecting the
composition of the sample. Triplicates of each of the samples were analysed using
PCA-CVA to minimise the variance between the triplicates (within-group
variance) and to maximise the variance between the individual time-point samples
(between-group variance). The results are shown in Fig. 8 .
Finally, for an online application of PS inferential estimation and control, as
proposed in Fig. 9 , methods capable of assigning online data into individual PSs
without relying upon the time-consuming and often un-economical offline analyses
have to be developed.