Attention module incorporated transfer learning empowered deep learning-based models for classification of phenotypically similar tropical cattle breeds (Bos indicus).

Accurate breed identification in dairy cattle is essential for optimizing herd management and improving genetic standards. A smart method for correctly identifying phenotypically similar breeds can empower farmers to enhance herd productivity. A convolutional neural network (CNN) based model was developed for the identification of Sahiwal and Red Sindhi cows. To increase the classification accuracy, first, cows's pixels were segmented from the background using CNN model. Using this segmented image, a masked image was produced by retaining cows' pixels from the original image while eliminating the background. To improve the classification accuracy, models were trained on four different images of each cow: front view, side view, grayscale front view, and grayscale side view. The masked images of these views were fed to the multi-input CNN model which predicts the class of input images. The segmentation model achieved intersection-over-union (IoU) and F1-score values of 81.75% and 85.26%, respectively with an inference time of 296 ms. For the classification task, multiple variants of MobileNet and EfficientNet models were used as the backbone along with pre-trained weights. The MobileNet model achieved 80.0% accuracy for both breeds, while MobileNetV2 and MobileNetV3 reached 82.0% accuracy. CNN models with EfficientNet as backbones outperformed MobileNet models, with accuracy ranging from 84.0% to 86.0%. The F1-scores for these models were found to be above 83.0%, indicating effective breed classification with fewer false positives and negatives. Thus, the present study demonstrates that deep learning models can be used effectively to identify phenotypically similar-looking cattle breeds. To accurately identify zebu breeds, this study will reduce the dependence of farmers on experts.

Allelic diversity at BoLA DRB3 locus and association with predisposition to clinical mastitis in indicus and crossbred cattle.

Bovine lymphocyte antigen (BoLA) DRB3 locus in healthy and mastitis affected cattle has been genotyped by a polymerase chain reaction and restriction fragment length polymorphisms (PCR-RLFP) using RsaI restriction enzyme, followed by sequencing. In 130 farm animals, 25 BoLA DRB3 alleles have been detected by PCR-RFLP. Three distinct allelic patterns significantly associated with mastitis in Karan Fries crossbred and Sahiwal indicus cattle have been identified, whereas, four other allelic patterns were significantly high in frequency among healthy animals. Sequencing of RFLP genotypes revealed 25 and 47 alleles among healthy Sahiwal and Karan Fries, respectively, while 17 and 38 patterns observed in mastitis affected Sahiwal and Karan Fries animals, respectively. From Tajima's D-test of neutrality, it was concluded that alleles associated with mastitis were expanding in the population, whereas those of healthy were under contraction. Phylogenetic analysis carried out to delineate the evolutionary relationship of the farm and field animals at DRB3 locus, differentiating allelic patterns into six different clusters. Among the phylogenetic lineages, five patterns DRB3*028:01, DRB3*011:03, DRB3*031:01, DRB3*001:01 and DRB3*043:01, were previously reported, whereas one novel allelic variant was observed in indicus and crossbred cattle. This information will help in further exploring the association between BoLA-DRB3 genetic diversity and disease resistance in distinct cattle breeds, important in designing breeding strategies for increasing the distribution of favorable alleles.

Genetic analysis of milk and milk composition traits in Murrah buffaloes using Bayesian inference.

Accurate and unbiased assessment of genetic parameters of milk and milk composition traits play an important role in formulating breeding program for genetic improvement of Murrah buffaloes. In this study, data spread over 28 years were utilized to estimate genetic parameters of traits viz., 305 d milk yield (305MY), 305 d fat yield (305FY), 305 d solid not fat yield (305SNFY), milk fat percentage (fat%) and solid not fat percentage (SNF) percentage (SNF%) in Murrah buffaloes kept at ICAR-National Dairy Research Institute, Karnal. Bayesian multiple-trait analysis was done using animal model and Gibbs sampling to estimate (co)variance components. Posterior means of heritability and posterior standard deviation for 305MY, 305FY, 305SNFY, fat% and SNF% were 0.18 ± 0.05, 0.17 ± 0.05, 0.18 ± 0.05, 0.07 ± 0.03 and 0.15 ± 0.06 and posterior means of repeatability estimates along with posterior standard deviation for corresponding traits were 0.33 ± 0.04, 0.32 ± 0.04, 0.33 ± 0.04, 0.14 ± 0.02 and 0.30 ± 0.04, respectively. Estimates of genetic correlation varied from -0.080 (305MY and fat %) to 0.999 (305MY and 305SNFY). Permanent environmental correlations varied from -0.060 (305MY and SNF%) to 0.999 (305FY and 305SNFY). This study indicated that all considered traits except fat% have ample genetic variability which can be exploited for selection and genetic improvement of Murrah buffaloes.

Casein (CSN) gene variants and parity affect the milk protein traits in crossbred (Bos taurus x Bos indicus) cows in sub-tropical climate.

Genotypes at four casein (CSN) loci-A26181G of CSN1S1, C6227T of CSN1S2, A8101C of CSN2, and A13104C of CSN3-along with non-genetic factors were studied for their effects on various milk protein traits in 100 crossbred cows with major inheritance of Holstein Frisian (Bos taurus) and Tharparkar (Bos indicus). Results demonstrated the presence of all CSN genotypes with a predominance of heterozygotes. At CSN2 (A8101C; His67Pro) locus, the A2 allele, desirable for human health, was present in 62% as heterozygous and 29% in homozygous condition. Among non-genetic factors, parity of the cows had a significant influence on the milk protein traits in these crossbreds. The genotypes at CSN1S1, CSN2, and CSN3 loci were found to influence (p<0.05 to 0.01) the casein and whey protein yields and composition traits. The casein index and total milk yield were most influenced by the CSN1S2 locus. The AA (A1 milk) genotype of CSN2 had significantly higher yields and percentages of casein and whey proteins. Positive influence of CC genotype of CSNS3 on milk proteins of was observed similar to Bos taurus cows; however, such influence of AA genotype of CSN2 locus may be distinctive to the crossbred cows maintained in subtropical condition. Overall, the results revealed the diverse effects of CSN genotypes on milk proteins in crossbred cattle.

Identification of Internal Reference Genes in Peripheral Blood Mononuclear Cells of Cattle Populations Adapted to Hot Arid Normoxia and Cold Arid Hypoxia Environments.

To estimate gene expression in a reliable manner, quantitative real-time polymerase chain reaction data require normalisation using a panel of stably expressed reference genes (RGs). To date, information on an appropriate panel of RGs in cattle populations reared at cold arid high-altitude hypoxia and hot arid tropical normoxia environments is not available. Therefore, the present study was carried out to identify a panel of stably expressed RGs from 10 candidate genes (GAPDH, RPL4, EEF1A1, RPS9, HPRT1, UXT, HMBS, B2M, RPS15, and ACTB) in peripheral blood mononuclear cells (PBMCs) of cattle populations reared at cold arid high-altitude hypoxia and hot arid normoxia environments. Four different statistical algorithms: geNorm, NormFinder, BestKeeper, and RefFinder were used to assess the stability of these genes. A total of 30 blood samples were collected: six adult heifers each of Ladakhi (LAC) and Holstein Frisian crosses (HFX) and 4 Jersey (JYC) cows from cold arid high-altitude hypoxia environments (group I) and five adult heifers each of Sahiwal (SAC), Karan Fries (KFC), and Holstein Friesian (HFC) cows from hot arid normoxia environments (group II). Combined analysis of group I and group II resulted in identification of a panel of RGs like RPS9, RPS15, and GAPDH that could act as a useful resource to unravel the accurate transcriptional profile of PBMCs from diverse cattle populations adapted to distinct altitudes.

Effect of Dickkopf-1 and colony stimulating factor-2 on the developmental competence, quality, gene expression and live birth rate of buffalo (Bubalus bubalis) embryos produced by hand-made cloning.

A functional canonical WNT signaling pathway exists in preimplantation embryos and inhibits embryonic development. Recent studies suggest that this pathway is over-expressed in nuclear transferred (NT), compared to IVF embryos. The present study investigated the effects of Dickkopf-1 (DKK1), an inhibitor of canonical WNT signaling pathway and colony stimulating factor-2 (CSF2), an embryokine, on the developmental competence, quality, gene expression and live birth rate of NT buffalo embryos produced by Hand-made cloning (HMC). Following supplementation of the in vitro culture medium on day 5 with DKK1 (100 ng/mL), CSF2 (10 ng/mL), DKK1+CSF2 or no supplementation (control), the blastocyst rate was higher (P < 0.05) with DKK1 and DKK1+CSF2 (42.6 ± 1.4% and 46.6 ± 0.9%, respectively) than with CSF2 or controls (40.6 ± 1.3% and 39.0 ± 1.3%, respectively). The apoptotic index of the blastocysts was lower (P < 0.05) for DKK1, CSF2 and DKK1+CSF2 groups (3.44 ± 0.14, 3.39 ± 0.11 and 3.11 ± 0.22, respectively) compared to controls (6.64 ± 0.25), and was similar to that of the IVF blastocysts (3.67 ± 0.18). Although the total cell number was similar for the DKK1, CSF2, DKK1+CSF2 and control groups (200.4 ± 3.05, 196.4 ± 3.73, 204.7 ± 3.71 and 205 ± 4.03, respectively), the inner cell mass:trophectoderm cell number ratio of DKK1, CSF2 and DKK1+CSF2 groups (0.21 ± 0.01, 0.17 ± 0.01 and 0.22 ± 0.02, respectively) was higher (P < 0.05) than controls (0.13 ± 0.01) and was similar to that of IVF blastocysts (0.19 ± 0.01). Treatment with DKK1 or CSF2 or both increased (P < 0.05) the expression level of OCT4, NANOG,SOX2, GATA6, BCL2, PTEN, P53, FGF4, GLUT1 and IFN-τ, and decreased that of C-MYC, CDX2, CASPASE, DNMT3a, TCF7 and LEF1 in blastocysts, compared to controls. Transfer of DKK1-treated embryos to 13 recipients resulted in 4 pregnancies (30.8%; 2 live births, one abortion and one currently at 9 months of pregnancy) whereas, transfer of DKK1+CSF2-treated embryos to 16 recipients, resulted in 4 pregnancies (25.0%), all of which resulted in live births. No pregnancy was obtained after transfer of control and CSF-treated embryos to 12 and 16 recipients, respectively. These results suggest that DKK1 treatment of NT embryos increases the blastocyst, conception and live birth rate, and improves their quality whereas, CSF2 treatment, does not affect the blastocyst, conception and live birth rate despite improvement in embryo quality.

Effect of supplementation with rumen-protected choline and green tea extract on production performance of transition Karan Fries cows.

AIM: The main objective of this study was to estimate the effect of supplementation of rumen-protected choline (RPC) and green tea extract (GTE) on production parameters in transition Karan Fries (KF) cows. MATERIALS AND METHODS: The present experiment was carried out on 32 pregnant KF cows. In the control group, cows were fed basal diet. In T1, each cow was fed RPC (55 g/day), in T2 - GTE (3 g/d), and in T3- RPC + GTE (55+3) g/day along with basal diet. The duration of the experiment was 30 days before calving to 60 days after parturition. RESULTS: Feeding of both RPC and GTE significantly increased milk yield (p≤0.01), 4% fat corrected milk yield (p≤0.01), milk fat (p≤0.01), and total solid content (p≤0.05) than control. There was no significant difference (p≥0.05) in milk protein, lactose, and solids not fat (SNF) content among the groups. CONCLUSION: Supplementing RPC and GTE in combination improved milk yield and fat content of the milk without altering protein, lactose content of the milk in transition KF cows.

Global proteomic analysis of water buffalo (Bubalus bubalis) saliva at different stages of estrous cycle using high throughput mass spectrometry.

Accurate and efficient detection of estrus is one of the major constraints for exploitation of the production potential of buffalo owing to its poor manifestation of estrus signs, seasonal differences in expression and higher incidences of silent estrus (29%). The current study focused on identification of estrus specific candidate proteins in saliva of buffaloes. Estrus was detected based on behavioral signs in response to the teaser and changes in reproductive organs and confirmed by per-rectal examination, trans-rectal USG of reproductive organs, cervico-vaginal mucus characteristics and blood serum progesterone estimation. Day of onset of estrus was considered as day 0 and day -3, +3, +10 were considered as proestrus, metestrus and diestrus stage of the estrous cycle respectively. A total of 19 animals and their 38 estrous cycles (two from each) were included in this study. Saliva was collected from these animals during different stages of estrous cycle. Out of these, 08 animals were selected for global proteome analysis of saliva using in-solution digestion and nano-LC-MS/MS. A total of 275, 371, 304 and 565 proteins were identified with ≥2 peptides during proestrus, estrus, metestrus and diestrus stages of estrous cycle. Among the identified proteins 31, 62, 32 and 104 proteins were found specific to proestrus, estrus, metestrus and diestrus stage of the estrous cycle. Few salivary proteins such as Cullin-associated NEDD8-dissociated protein 1, Heat shock 70 kDa protein 1A, 17-beta-hydroxysteroid dehydrogenase type 1, Inhibin beta A chain, testin were identified as estrus specific and are important for estrus physiology. Taken together, these estrus specific proteins could be considered as the candidate biomarker for detection and confirmation of estrus in buffalo after thorough validation.

Biochemical analysis of uterine fluid for identification of indicators for subclinical endometritis in the water buffalo (Bubalus bubalis).

Alterations in biochemical constituents of uterine fluid have been suggested for diagnosis of subclinical uterine infection in the bovine. This study was undertaken to investigate whether uterine fluid biomolecules could act as tool for diagnosis of subclinical endometritis in the buffalo. Uterine fluid samples from normal (n = 22) and subclinical endometritis (n = 18; diagnosed based on uterine cytology)-affected buffaloes were subjected to biochemical analysis. Among the different biochemical constituents estimated, urea, urea N, cholesterol, total bilirubin and alkaline phosphatase (ALP) concentrations were significantly (p < 0.05) higher in uterine fluid obtained from subclinical endometritis-affected buffaloes. The extent of difference between normal and subclinical endometritis-affected buffaloes was highest in ALP (69%) followed by cholesterol (55%), bilirubin (48%), urea (30%) and urea N (30%) concentrations. Receiver operating characteristic (ROC) analysis indicated that the likelihood ratio (LR) was 3.63 for urea, indicating that buffaloes having less than the threshold concentration (47.5 mg/dl) of urea in their uterine fluid were at 3.6 times more risk to be affected with SE. The LRs for urea N, cholesterol, ALP and bilirubin were 2.33, 2.54, 2.12 and 1.65, respectively. It was concluded that ALP, urea, urea N and cholesterol concentrations in uterine fluid may serve as an aid for diagnosing subclinical endometritis in the buffalo.

Efficiency of uterine fluid cytology in the diagnosis of subclinical endometritis in the water buffalo (Bubalus bubalis).

This study compared endometrial cytology vis-a-vis uterine fluid cytology for assessment of uterine health in clinically normal and subclinical endometritis (SE)-affected buffaloes. Uterine fluid samples and endometrial samples were collected from the buffaloes (n = 38) at oestrus using blue sheath and cytobrush, respectively. The smears were stained with Field stain for 3 minutes, and a minimum of 400 cells were counted in each smear for determination of the percentage of polymorphonuclear (PMN) leucocyte. The incidence of subclinical endometritis, based on the cytobrush cytology, was 23.08%. The correlation between cytobrush cytology with uterine fluid cytology was positive and significant (r = .37; p = .02). The ratio of PMN leucocyte in cytobrush cytology to uterine fluid cytology was 1:2.4. ROC analysis revealed that the threshold value of 6.16% PMN leucocyte in uterine fluid cytology showed a diagnostic sensitivity and specificity of 100% in differentiating normal from SE-affected buffaloes. In conclusion, collection of uterine fluid was easier compared to collection of endometrial samples using cytobrush and the percentage of PMN leucocyte in uterine fluid cytology can be used as a tool for diagnosis of subclinical endometritis in buffaloes.

Study of short lactation in Sahiwal cattle at organized farm.

AIM: The aim was to study the associated factors and extent of short lactations in Sahiwal cattle maintained under organized herd. MATERIALS AND METHODS: The present study was conducted on Sahiwal cattle (n=530), utilizing 1724 lactation records with respect to lactation length (LL), spread over a period of 15 years (1997-2011), maintained at Livestock Research Center, National Dairy Research Institute, Karnal. Observations of LL were analyzed by descriptive statistical analysis in order to know the extent of short lactation of animals in the herd. Paternal Half sib method was used to estimate the genetic parameters, i.e., heritability, genetic, and phenotypic correlation. The influence of various non-genetic factors (season of calving, the period of calving, parity, type of calving, and season of drying) on LL was studied by least squares analysis of variance technique. RESULTS: The least squares means for LL was found to be 215.83±3.08 days. Only 32.48% of total lactation records were fell in the range of 251-350 days of LL, while more than three-fourth (76%) of total observations were failed to reach the standard level of 305 milking days. LL class ranges from 251 to 300 days accommodated maximum number of observations (19.2%). The heritability estimate of LL was 0.22±0.07. Positive correlations were found between LL and service period, LL and 305 or less days milk yield, LL and calving interval; whereas dry period was negatively correlated with the LL. The least squares analysis had shown that LL was significantly (p<0.01) influenced by the period of calving, type of calving, and season of drying. Significantly higher LL (276.50±7.21 days) was found in animals calved in the first period than those calved in other periods. The cows dried during summer season had the shortest LL (188.48±7.68 days) as compared to other seasons. CONCLUSION: Present findings regarding short lactations occurrence may be alarming for the indigenous herd, demanding comprehensive study with the larger data set. Since LL was influenced by various environmental factors suggesting better managerial tools, besides special attention on the milch animals going to dry during the summer season.