Integrating metabolome dynamics and process data to guide cell line selection in biopharmaceutical process development

Barberi, Gianmarco; Benedetti, Antonio; Diaz-Fernandez, Paloma; Sévin, Daniel C.; Vappiani, Johanna; Finka, Gary; Bezzo, Fabrizio; Barolo, Massimiliano; Facco, Pierantonio

doi:10.1016/j.ymben.2022.03.015

“…Unsupervised techniques are predominantly useful in the validation of sample profiles, identification of subpopulations, detection of biological patterns, and integration of multi-omic data (Box 3). Principal component analysis (PCA) is commonly applied as quality control to validate sample similarities and to detect obvious technical factors such as batch effects [39][40][41][42][43][44][45], while k-means or hierarchical clustering can facilitate the identification of subject subpopulations or feature groups that exhibit similar behaviors [42,45,46]. Independent component analysis (ICA) and Markov clustering can identify biologically meaningful interactions between molecular species as demonstrated in E.coli [47] and human[48] omic data.…”

Section: Data-driven Inference Modelsmentioning

confidence: 99%

From Observational to Actionable: Rethinking Omics in Biologics Production

Masson¹,

Karottki²,

Tat³

et al. 2023

Preprint

0

View full text Add to dashboard Cite

As the era of omics continues to expand, omics-based experiments have become popular in industrial biotechnology. They promise deeper biological understanding, which may be leveraged to develop novel solutions to bioprocess optimization and cell line engineering strategies. Despite this expansion, naivety about what omic data offers and how to best handle such data can challenge the extraction of actionable value. However, the value of omic experiments in biotechnology research and development can be maximized with deliberate application of omic approaches and forethought about analysis techniques. Here we describe important considerations when designing and implementing omic-based experiments, and discuss how systems biology analysis strategies can enhance efforts to obtain actionable insights in biomanufacturing.

show abstract

“…Unsupervised techniques are predominantly useful in the validation of sample profiles, identification of subpopulations, detection of biological patterns, and integration of multi-omic data (Box 3). Principal component analysis (PCA) is commonly applied as quality control to validate sample similarities and to detect obvious technical factors such as batch effects [39][40][41][42][43][44][45], while k-means or hierarchical clustering can facilitate the identification of subject subpopulations or feature groups that exhibit similar behaviors [42,45,46]. Independent component analysis (ICA) and Markov clustering can identify biologically meaningful interactions between molecular species as demonstrated in E.coli [47] and human[48] omic data.…”

Section: Data-driven Inference Modelsmentioning

confidence: 99%

From observational to actionable: rethinking omics in biologics production

Masson

¹

,

Karottki

²

,

Tat

³

et al. 2023

Trends in Biotechnology

2

0

View full text Add to dashboard Cite

“…To improve PLS estimations, variable selection [9,44] is used, in such a way as to identify and retain in the model only the variables with the largest information content on the mAbs titer and exclude the other variables. Variable importance is assessed through the variable importance in projection (VIP) [45] index:…”

Section: Multivariate Predictive Modelingmentioning

confidence: 99%

“…Following the wave of digitalization in Industry 4.0, large amounts of data (e.g., culture variables from high throughput technologies [7], and omics data such as transcriptomics [8] or metabolomics [9]) are usually collected from all the stages of the scale-up. The wealth of information contained in the experimental data can be extracted to support the mAbs development through machine learning [9,10].…”

Section: Introductionmentioning

confidence: 99%

“…Following the wave of digitalization in Industry 4.0, large amounts of data (e.g., culture variables from high throughput technologies [7], and omics data such as transcriptomics [8] or metabolomics [9]) are usually collected from all the stages of the scale-up. The wealth of information contained in the experimental data can be extracted to support the mAbs development through machine learning [9,10]. In particular, different data-driven techniques were demonstrated to be effective to: (i) understand the similarity among bioreactors at different scales and improve the similarity between scales in the scaled-down [11]; (ii) predict the mAbs concentration at harvest allowing to identify the parameters that promote or suppress production [12]; (iii) estimate the mAbs quality and interpret the relationship between process and product when coupled with genetic algorithms [13]; and (iv) capture very complex biological relationships through neural networks coupled with first principles models of the culture environment and accurately predict the mAbs quality attributes [14].…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Data Augmentation to Support Biopharmaceutical Process Development through Digital Models—A Proof of Concept

Botton

¹

,

Barberi

²

,

Facco

³

2022

Self Cite

View full text Add to dashboard Cite

In recent years, monoclonal antibodies (mAbs) are gaining a wide market share as the most impactful bioproducts. The development of mAbs requires extensive experimental campaigns which may last several years and cost billions of dollars. Following the paradigm of Industry 4.0 digitalization, data-driven methodologies are now used to accelerate the development of new biopharmaceutical products. For instance, predictive models can be built to forecast the productivity of the cell lines in the culture in such a way as to anticipate the identification of the cell lines to be progressed in the scale-up exercise. However, the number of experiments that can be performed decreases dramatically as the process scale increases, due to the resources required for each experimental run. This limits the availability of experimental data and, accordingly, the applicability of data-driven methodologies to support the process development. To address this issue in this work we propose the use of digital models to generate in silico data and augment the amount of data available from real (i.e., in vivo) experimental runs, accordingly. In particular, we propose two strategies for in silico data generation to estimate the endpoint product titer in mAbs manufacturing: one based on a first principles model and one on a hybrid semi-parametric model. As a proof of concept, the effect of in silico data generation was investigated on a simulated biopharmaceutical process for the production of mAbs. We obtained very promising results: the digital model effectively supports the identification of high-productive cell lines (i.e., high mAb titer) even when a very low number of real experimental batches (two or three) is available.

show abstract

Integrating metabolome dynamics and process data to guide cell line selection in biopharmaceutical process development

Cited by 6 publications

References 36 publications

From Observational to Actionable: Rethinking Omics in Biologics Production

From Observational to Actionable: Rethinking Omics in Biologics Production

From observational to actionable: rethinking omics in biologics production

Data Augmentation to Support Biopharmaceutical Process Development through Digital Models—A Proof of Concept

Contact Info

Product

Resources

About