The search for SNPs and genes associated with the feed conversion ratio using entropy analysis

© Slovak University of Agriculture in Nitra Faculty of Agrobiology and Food Resources


Introduction
Feed conversion ratio is one of the most important traits in the broiler industry.Nutrition accounts for about 70% of the total costs of broiler production.Feed conversion ratio (FCR) is expressed as feed intake per body weight gain (Aggrey et al., 2010).It is well known that FCR is considerably correlated with other performance traits.However, a routine registration of this trait is relatively expensive.Hence, from practical perspective, identification of genome regions determining feed efficiency is very desirable.Recently, an information theory has been more and more applied to study the associations between SNPs and traits, mainly categorical ones (Borowska et al., 2014).Based on this approach the interactions between SNPs can be investigated.The aim of this study was to detect the chromosomal regions connected with FCR and point out genes which are connected with FCR in meat-type chicken.

Material and methods
Analysis covered 862 commercial broiler chickens genotyped using the 60 K Illumina iSelect chicken array.The genotyping revealed information about 57636 SNPs.Feed conversion ratio between 39-46 days was recorded for all genotyped individuals.More details of the experiment are given by Reyer et al. (2015).After data editing, 855 individuals and 28 chromosomes (except chromosomes 16 and W) were analysed.The trait values were classified according to quartiles and ranged from 1-4.
The data editing based on the following criteria: minor allele frequency >5% and call rate >95%.Finally, 42770 SNPs were included.
Entropy analysis was employed to study important genome regions contributing to FCR.The following parameters were estimated (Borowska et al., 2014): entropy coefficient, conditional entropy, portion of information and mutual information.
Chromosomal regions were indicated based on the highest mutual information.Higher mutual information between two SNPs means that one SNP is associated with the other.SNPs located in intragenic regions were preferred for the last part of the analysis.The location of SNPs and genes were verified using the Ensembl genome database (EMBL-EBI, 2016) and the National Center for Biotechnology Information (NCBI) databases -db SNP and Gene (NCBI, 2016a, b).

Results and discussion
Two main regions on chromosomes 1 and Z were detected for FCR by entropy analyses, depend on the highest portion on information assigned (entropy of the trait -conditional entropy; ET-EC) and highest mutual information (Figure 1).
Based on the assumption of 'known status' in chicken and with SNP located inside the gene, five genes were pointed out: CACNA2D1, GNAI1, JAK2, SEMA3A, SEMA3D.CACNA2D1 is involved in the development of the skin epithelium in chicken embryos (Chang et al., 2015).GNAI1 is connected with many signalling pathways.The influence of JAK2 gene on growth and reproduction trait was confirmed in meat-type chicken (Liu et al., 2010).The expression of SEMA3A was observed in central nervous system, typically in gray matter of spinal cord in chicken embryos (Fu et al. 2000).SEMA3D is also connected with the development of nervous system e.g. in brain.Moreover, it is participating in the development of lens and nasal placodes, the heart, and in the ectoderm of limbs (Bao and Jin, 2006;Jin et al. 2006).Additionally, the expressed proteins of GNAI1, JAK2 and SEMA3A were detected in the gastrointestinal tract (liver, gall bladder, stomach, small intestine etc.) in humans (Pontén et al., 2008).

Conclusions
The entropy analysis allows to search for associations between SNPs (in the context of some functional genes) and feed efficiency.These genes (including important SNPs) are connected with nervous system and expressed also in gastrointestinal tract.Hence, their considerable role in chicken feed conversion is suggested.

Figure 1
Location of 1% and 5% the most infomative SNPs for feed conversion ratio