The recipient of a PhD in biology from Yale University, Thomas Brutnell is a co-founder of Benson Hill Biosystems. With over two decades of experience in plant molecular biology, Thomas Brutnell has advised industry scientists as well as academic scientists on effective plant breeding techniques that improve selection.
A field of modern biotechnology, plant comparative genomic analysis focuses on predicting functions of various unknown genes in plant genomes by analyzing the similarities and differences among species. Developing the correct strategies and tools to manage information has been a key requirement for the success of the analysis. Recently, various approaches in systems biology and bioinformatics have helped to address this need. However, in this nascent field new disruptive technologies pose challenges as existing platforms must be tweaked to accommodate new data channels. A good example is single cell sequencing in which transcripts are grouped based on co-expression analysis and cell types are defined using marker gene expression to seed the analysis. Thus, ideally, co-expression network data would be integrated with in situ hybridization data, but currently no database has this capability. The challenges are multi-fold. In this example, most in situ data is not digitized, it is usually presented as figures in the body of a manuscript or in supplemental materials. Furthermore, the author often is often subjective in the interpretation of the data; focussing on the points they want to illustrate and often not comprehensive in describing all of the possible expression domains. Thus, moving from in situ data from one organism to infer function in another is quite problematic.
In addition to analytical tools, many powerful online databases have been created to manage and store bulk genomic data obtained through comparative analysis. GreenPhylDB, PLAZA, Phytozome, and PlantsDB are some examples of community databases. However, all of these platforms struggle to stay current with new genomic datasets as the pace of genome sequencing is intensifying while the funding and personnel to maintain these databases is not. To keep pace it will be essential for curators to adopt AI and machine learning tools to help integrate and automate data processing.
A field of modern biotechnology, plant comparative genomic analysis focuses on predicting functions of various unknown genes in plant genomes by analyzing the similarities and differences among species. Developing the correct strategies and tools to manage information has been a key requirement for the success of the analysis. Recently, various approaches in systems biology and bioinformatics have helped to address this need. However, in this nascent field new disruptive technologies pose challenges as existing platforms must be tweaked to accommodate new data channels. A good example is single cell sequencing in which transcripts are grouped based on co-expression analysis and cell types are defined using marker gene expression to seed the analysis. Thus, ideally, co-expression network data would be integrated with in situ hybridization data, but currently no database has this capability. The challenges are multi-fold. In this example, most in situ data is not digitized, it is usually presented as figures in the body of a manuscript or in supplemental materials. Furthermore, the author often is often subjective in the interpretation of the data; focussing on the points they want to illustrate and often not comprehensive in describing all of the possible expression domains. Thus, moving from in situ data from one organism to infer function in another is quite problematic.
In addition to analytical tools, many powerful online databases have been created to manage and store bulk genomic data obtained through comparative analysis. GreenPhylDB, PLAZA, Phytozome, and PlantsDB are some examples of community databases. However, all of these platforms struggle to stay current with new genomic datasets as the pace of genome sequencing is intensifying while the funding and personnel to maintain these databases is not. To keep pace it will be essential for curators to adopt AI and machine learning tools to help integrate and automate data processing.
