Abstract
In this study, sequencing of the 16S rRNA gene targeting the V4-V6 regions was conducted to assess the cecal microbial alterations in response to dietary supplementation with a yeast derived mannan rich fraction (MRF) in standard commercial broiler production settings across four separate broiler trials. The resulting data was analysed to identify consistent changes in the bacterial community structure of the broiler cecum in response to MRF supplementation. Subsequently, the datasets from each individual trial were pooled and analysed for differences between control and MRF supplemented diets at day 35 posthatch. The results from this analysis showed that Phylum Firmicutes was decreased and Phylum Bacteroidetes was increased across all four trials at day 35 posthatch when compared to the control. An extension of the random forest bioinformatics approach to discover a highly relevant set of microbial operational taxonomic units (OTUs) which are indicative of MRF supplementation in the broiler cecum was then used. This approach has enabled the identification of a novel set of yeast-mannan sensitive bacterial OTUs in the cecal microbiome. This information will be helpful in developing potential future nutritional strategies and will be favourable to the poultry industry.
Highlights
The health and nutritional state of the broiler is largely interlinked with the gastrointestinal (GI) microbiome
We sought to merge the existing knowledge of these two published trials with two additional broiler trials to identify the dominant changes in the principal taxonomic categories and to assess if we could identify a consistent set of mannan rich fraction (MRF) sensitive bacterial OTUs in the broiler cecum using an extended RFM of classification
Results indicated that cecal bacterial community composition (BCC) was significantly altered as a result of dietary MRF in both trials
Summary
The health and nutritional state of the broiler is largely interlinked with the gastrointestinal (GI) microbiome. Advances in bioinformatic techniques are making it possible to identify differences in these distinct but low abundant microbial taxa and implicate them with a physiological outcome[16,17] These recent advances in biological data acquisition and sequencing technologies are enabling identification of thousands, sometimes millions, of features for a sample[18]. An Extended Conditional Inference Forest (ECIF) approach, developed by our team to remove variables less relevant than random predictors and used successfully on multiple applications, is a novel method for better feature selection and ranking accuracy, which are critical tasks for addressing the challenges outlined above It was based on the Boruta Package[20]. This method has been used by the team successfully for other research efforts and was applied in this analysis[21,22]
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
