28.05.2019 - 17:00
Transcriptional regulatory networks are at the core of plant responses to developmental, biotic, and abiotic cues. Yet, the majority of transcription factors remain without an associated function in most species. We have deployed random forest based machine learning (ML) to predict target genes of all transcription factors of a species based on publicly available RNA-seq data. ML enables us to annotate their functions based on their predicted targets. We test the functional annotation against database standards, and test the association of targets to transcription factors using RNA-seq, overexpressors and mutants. Networks in five species are currently available and enable evolutionary comparisons between species and within species for those who are auto- or allopolyploid. The gene regulatory networks are used to explore transcription factor binding motifs in their predicted target genes.
In the talk I will present strengths and limitations of ML based analyses in gene regulatory network analyses.
For more information see:
If you would like to meet Andrea Bräutigam personally, please contact Marcel Quint (email@example.com).
Redner: Andrea Bräutigam
Ort: Lecture Hall E.02
Gastgeber: Marcel Quint (firstname.lastname@example.org)
Anschrift: Bielefeld University