Gastrointestinal Spatiotemporal mRNA Expression of Ghrelin vs Growth Hormone Receptor and New Growth Yield Machine Learning Model Based on Perturbation Theory

Abstract

The management of ruminant growth yield has economic importance. The current work presents a study of the spatiotemporal dynamic expression of Ghrelin and GHR at mRNA levels throughout the gastrointestinal tract (GIT) of kid goats under housing and grazing systems. The experiments show that the feeding system and age affected the expression of either Ghrelin or GHR with different mechanisms. Furthermore, the experimental data are used to build new Machine Learning models based on the Perturbation Theory, which can predict the effects of perturbations of Ghrelin and GHR mRNA expression on the growth yield. The models consider eight longitudinal GIT segments (rumen, abomasum, duodenum, jejunum, ileum, cecum, colon and rectum), seven time points (0, 7, 14, 28, 42, 56 and 70 d) and two feeding systems (Supplemental and Grazing feeding) as perturbations from the expected values of the growth yield. The best regression model was obtained using Random Forest, with the coefficient of determination R2 of 0.781 for the test subset. The current results indicate that the non-linear regression model can accurately predict the growth yield and the key nodes during gastrointestinal development, which is helpful to optimize the feeding management strategies in ruminant production system.

Experimental Study and ANN Dual-Time Scale Perturbation Model of Electrokinetic Properties of Microbiota

“…This idea was widely used to solve many practical problems, such as the fatty acid distribution (Liu et al, 2015), goat growth yield via mRNA expression of ghrelin receptor and growth hormone receptor (Ran et al, 2016), carbon nanotubes (González-Durruthy et al, 2017), and drug-lymphocyte interactome networks (Tenorio-Borroto et al, 2016), etc.…”

“…The first type is the variable < ξ>, which is the Expected Measurement (EM) component used to account for the expected value of the output property marked as ξ (e) , where the subscript “e” is the expected measurement. This theoretical notion has been widely used in our previous works (Ran et al, 2016 ). The other type refers to the Box-Jenkins Operators (or perturbation values) ΔV ϕ (t k ), t k representing the time scale t 1 and t 2 , while V ϕ is V q or V f , respectively.…”

Prediction of Antimalarial Drug-Decorated Nanoparticle Delivery Systems with Random Forest Models

“…Moreover, a brand new method for data fusion in nanotechnology, bio-molecular sciences, chemistry and big data analysis has been proposed in different works: it integrates Perturbation Theory (PT) and Machine Learning (ML) [ 3 , 4 , 5 , 6 , 7 , 8 , 9 , 10 , 11 , 12 , 13 ], using distinct PT operators to analyze changes in the varied non-structural and structural conditions of a test at once (PTML). A few of these PT operators represent the generalization of a classic cheminformatics approach introduced by Corwin Hansch [ 14 ].…”

Improvement of Epitope Prediction Using Peptide Sequence Descriptors and Machine Learning

“…The computational methods to find a mathematical model to predict peptide function using its molecular information are numerous, from linear to non-linear algorithms. The qualitative structure–activity/property relationships (QSAR/QSPR) [10,11] represent very useful methods that are able to predict molecular activity or property using molecular descriptors. In order to find the mathematical function/algorithm that is able to map the inputs (molecule descriptors/characteristics) to the output (molecule biological activity), linear and non-linear Machine Learning (ML) methods are used.…”