Generation Interval Research Articles

Synergizing edge computing and blockchain for cyber‐physical systems

SummaryAs a foundational technology for managing decentralized systems like smart grids and healthcare systems, blockchain is attracting a lot of interest. However, owing to excessive resource requirements and low scalability with frequent‐intensive transactions, its use in resource‐constrained mobile devices is restricted. To let mobile devices offload processing tasks to cloud resources, edge computing can be used. For scalability and secure transactions, the integration with edge computing guarantees consistent access, distributed computing, and non‐compromised storage. To accomplish effective integration, critical challenges such as reliability, adaptability, and resource planning must be resolved. Despite these efforts, further study is still needed to address issues with confidentiality, flexibility, and reliability to develop a workable, secure decentralized data storage system. This study utilizes edge computing to build an Internet of Things (IoT) architecture that satisfies the safety and scalability necessary criteria for integration along with the integration of peer‐to‐peer and blockchain technologies. Already‐existing blockchain and related technologies have also been investigated to suggest solutions that address issues like secrecy, reliability, and scalability in order to successfully incorporate blockchain in IoT systems. Our experimental findings indicate that achieving low rates of stale blocks and promoting decentralization requires a meticulous selection of block sizes and generation intervals. It is essential to maintain block creation intervals as short as practically possible, and utilizing blocks with a size below 1 MB is advisable.

A Bayesian approach using spatiotemporal features for suitable next hop selection in opportunistic networks

SummaryOpportunistic network (OppNet) belongs to the category of Mobile Ad‐hoc Networks (MANETs), a kind of Delay Tolerant Network (DTN), where the wireless nodes are completely mobile and the data transmission routes are dynamic. The major challenge in developing a routing model for such a network is the unpredictable nature of the movement of the nodes. In this paper, a spatiotemporal prediction model based on human mobility pattern is proposed using Bayesian posterior probability (BPPR) where several clusters are identified within the network and the day and time duration of nodes visiting those clusters are recorded. The Bayesian posterior probability is then used to determine the probability of the neighbor node visiting the destination's cluster. If the calculated probability for that node is higher than a specified threshold, the packet will be forwarded. A comparison of the results obtained on simulation is made with benchmark models—Epidemic, Prophet, HBPR, EDR, NexT, and EBC, to name a few, where it is found that on average the proposed model outperforms the other models in terms of delivery probability, hop count and number of messages dropped by around 23.89%, 24.8%, 24.4%, 37%, 11%, and 42% respectively, with varying number of nodes, TTL, message generation interval, and buffer size. Similar improvements have been observed in terms of the other two metrics. In terms of overhead ratio, the proposed model outperforms Epidemic, Prophet, HBPR, NexT, and EBC. However, as the number of nodes and TTL are varied, BPPR performs better than NexT by around 9% and 12%, respectively, in terms of average latency.

Development of a non-programmable sperm cryopreservation technique in dwarf surfclams Mulinia lateralis–a potential model species for bivalve research

Sperm cryopreservation technique has been published in many farmed bivalve species. One of the key factors preventing its application in aquaculture and/or cryobanking is the knowledge gap on the performance of resultant progeny at late developmental stages and subsequent generations. An effective strategy to overcome these challenges is to use a model species with a short generation interval, such as the dwarf surfclam Mulinia lateralis (three months). This study evaluated the parameters key to the development of a non-programmable sperm cryopreservation technique in this species, with a D-stage larval rate similar to control being achieved when the sperm were cryopreserved under the conditions (cryoprotectant agent: 8% dimethyl sulfoxide; equilibration period: 10 min; rack height: 4 cm; thawing temperature: 60°C and sperm to egg ratio: 1100:1) optimized. This technique is the most common method applied in bivalve and the results from this study were all within the ranges published for other bivalve species, indicating this species would be an ideal bivalve model species for addressing cryopreservation evaluation issues that need a long extended time to collect data and/or challenging field experiments.

A hybrid model of CNN and LSTM autoencoder-based short-term PV power generation forecasting

AbstractSolar energy is one of the main renewable energies available to fulfill global clean energy targets. The main issue of solar energy like other renewable energies is its randomness and intermittency which affects power grids stability. As a solution for this issue, energy storage units could be used to store surplus energy and reuse it during low solar generation intervals. Also, in order to sustain stable power grid and better grid operation and energy storage management, photovoltaic (PV) power forecasting is inevitable. In this paper, new hybrid model based on deep learning techniques is proposed to predict short-term PV power generation. The proposed model incorporates convolutional neural network (CNN) and long short-term memory (LSTM) autoencoder network. The new model differentiates itself in accomplishing high prediction accuracy by extracting spatial features in time series via CNN layers and temporal features between the time series data through LSTM. The introduced model is tested on dataset of power generation from southern UK solar farm and the weather data corresponding to same location and time intervals; the forecasting performance of the suggested model is evaluated in metrics of root-mean-square error (RMSE) and mean absolute error (MAE). The used model is compared with different models from the literature either of pure type of network such as LSTM and gated recurrent unit (GRU) or hybrid combination of different networks like CNN-LSTM and CNN-GRU. The results show that proposed model provides enhanced results and reduces training time significantly compared to other competitive models, where the performance of the proposed model improved averagely by 5% to 25% in terms of RMSE and MAE performance metrics, and the execution time of training significantly reduced with almost 70% less compared to other models.

Genome-wide association study for stayability at different calvings in Nellore beef cattle

BackgroundingStayability, which may be defined as the probability of a cow remaining in the herd until a reference age or at a specific number of calvings, is usually measured late in the animal’s life. Thus, if used as selection criteria, it will increase the generation interval and consequently might decrease the annual genetic gain. Measuring stayability at an earlier age could be a reasonable strategy to avoid this problem. In this sense, a better understanding of the genetic architecture of this trait at different ages and/or at different calvings is important. This study was conducted to identify possible regions with major effects on stayability measured considering different numbers of calvings in Nellore cattle as well as pathways that can be involved in its expression throughout the female’s productive life.ResultsThe top 10 most important SNP windows explained, on average, 17.60% of the genetic additive variance for stayability, varying between 13.70% (at the eighth calving) and 21% (at the fifth calving). These SNP windows were located on 17 chromosomes (1, 2, 4, 6, 7, 8, 9, 10, 11, 12, 13, 14, 18, 19, 20, 27, and 28), and they harbored a total of 176 annotated genes. The functional analyses of these genes, in general, indicate that the expression of stayability from the second to the sixth calving is mainly affected by genetic factors related to reproductive performance, and nervous and immune systems. At the seventh and eighth calvings, genes and pathways related to animal health, such as density bone and cancer, might be more relevant.ConclusionOur results indicate that part of the target genomic regions in selecting for stayability at earlier ages (from the 2th to the 6th calving) would be different than selecting for this trait at later ages (7th and 8th calvings). While the expression of stayability at earlier ages appeared to be more influenced by genetic factors linked to reproductive performance together with an overall health/immunity, at later ages genetic factors related to an overall animal health gain relevance. These results support that selecting for stayability at earlier ages (perhaps at the second calving) could be applied, having practical implications in breeding programs since it could drastically reduce the generation interval, accelerating the genetic progress.

A short-term photovoltaic power interval forecasting method based on fuzzy granular computing and CNN-BiGRU

Abstract This paper presents a short-term PV power interval prediction method combining fuzzy information granulation and CNN-BiGRU model. First, historical data of PV power generation is processed using fuzzy information granulation to determine the interval range. Subsequently, a CNN-BiGRU model is constructed, where the CNN module extracts local features of the interval range and the BiGRU module captures temporal patterns. The interval range is then fed into the CNN-BiGRU model for training, which enables accurate prediction of short-term PV power generation intervals. Finally, the predicted power generation interval ranges are given. The experimental results show that this hybrid interval prediction method is first applied to the prediction of short-term PV power scenarios, which fully utilizes the advantages of information granularity to express the data intervals and CNN-BiGRU to learn the complex time series patterns, and realizes the accurate and reliable prediction of short-term PV power intervals. Compared with direct numerical prediction, the method provides an expression of prediction uncertainty, which provides an important reference for the operation decision of PV power plants.

Transmission characteristics and inactivated vaccine effectiveness against transmission of the SARS-CoV-2 Omicron BA.2 variant in Shenzhen, China.

We conducted a retrospective cohort study to evaluate the transmission risk of severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) Omicron BA.2 variant and the effectiveness of inactivated COVID-19 vaccine boosters in Shenzhen during a BA.2 outbreak period from 1 February to 21 April 2022. A total of 1,248 individuals were infected with the BA.2 variant, and 7,855 close contacts were carefully investigated. The risk factors for the high secondary attack rate of SARS-CoV-2 infection were household contacts [adjusted odds ratio (aOR): 1.748; 95% confidence interval (CI): 1.448, 2.110], younger individuals aged 0-17 years (aOR: 2.730; 95% CI: 2.118, 3.518), older persons aged ≥60 years (aOR: 1.342; 95% CI: 1.135, 1.588), women (aOR: 1.442; 95% CI: 1.210, 1.718), and the subjects exposed to the post-onset index cases (aOR: 8.546; 95% CI: 6.610, 11.050), respectively. Compared with the unvaccinated and partially vaccinated individuals, a relatively low risk of secondary attack was found for the individuals who received booster vaccination (aOR: 0.871; 95% CI: 0.761, 0.997). Moreover, a high transmission risk was found for the index cases aged ≥60 years (aOR: 1.359; 95% CI: 1.132, 1.632), whereas a relatively low transmission risk was observed for the index cases who received full vaccination (aOR: 0.642; 95% CI: 0.490, 0.841) and booster vaccination (aOR: 0.676; 95% CI: 0.594, 0.770). Compared with full vaccination, booster vaccination of inactivated COVID-19 vaccine showed an effectiveness of 24.0% (95% CI: 7.0%, 37.9%) against BA.2 transmission for the adults ≥18 years and 93.7% (95% CI: 72.4%, 98.6%) for the adults ≥60 years, whereas the effectiveness was 51.0% (95% CI: 21.9%, 69.3%) for the individuals of 14 days to 179 days after booster vaccination and 51.2% (95% CI: 37.5%, 61.9%) for the non-household contacts. The estimated mean values of the generation interval, serial interval, incubation period, latent period, and viral shedding period were 2.7 days, 3.2 days, 2.4 days, 2.1 days, and 17.9 days, respectively. In summary, our results confirmed that the main transmission route of Omicron BA.2 subvariant was household contact, and booster vaccination of the inactivated vaccines was relatively effective against BA.2 subvariant transmission in older people.

Genomic selection for target traits in the Australian lentil breeding program.

Genomic selection (GS) uses associations between markers and phenotypes to predict the breeding values of individuals. It can be applied early in the breeding cycle to reduce the cross-to-cross generation interval and thereby increase genetic gain per unit of time. The development of cost-effective, high-throughput genotyping platforms has revolutionized plant breeding programs by enabling the implementation of GS at the scale required to achieve impact. As a result, GS is becoming routine in plant breeding, even in minor crops such as pulses. Here we examined 2,081 breeding lines from Agriculture Victoria's national lentil breeding program for a range of target traits including grain yield, ascochyta blight resistance, botrytis grey mould resistance, salinity and boron stress tolerance, 100-grain weight, seed size index and protein content. A broad range of narrow-sense heritabilities was observed across these traits (0.24-0.66). Genomic prediction models were developed based on 64,781 genome-wide SNPs using Bayesian methodology and genomic estimated breeding values (GEBVs) were calculated. Forward cross-validation was applied to examine the prediction accuracy of GS for these targeted traits. The accuracy of GEBVs was consistently higher (0.34-0.83) than BLUP estimated breeding values (EBVs) (0.22-0.54), indicating a higher expected rate of genetic gain with GS. GS-led parental selection using early generation breeding materials also resulted in higher genetic gain compared to BLUP-based selection performed using later generation breeding lines. Our results show that implementing GS in lentil breeding will fast track the development of high-yielding cultivars with increased resistance to biotic and abiotic stresses, as well as improved seed quality traits.

SABO-ILSTSVR: a genomic prediction method based on improved least squares twin support vector regression.

In modern breeding practices, genomic prediction (GP) uses high-density single nucleotide polymorphisms (SNPs) markers to predict genomic estimated breeding values (GEBVs) for crucial phenotypes, thereby speeding up selection breeding process and shortening generation intervals. However, due to the characteristic of genotype data typically having far fewer sample numbers than SNPs markers, overfitting commonly arise during model training. To address this, the present study builds upon the Least Squares Twin Support Vector Regression (LSTSVR) model by incorporating a Lasso regularization term named ILSTSVR. Because of the complexity of parameter tuning for different datasets, subtraction average based optimizer (SABO) is further introduced to optimize ILSTSVR, and then obtain the GP model named SABO-ILSTSVR. Experiments conducted on four different crop datasets demonstrate that SABO-ILSTSVR outperforms or is equivalent in efficiency to widely-used genomic prediction methods. Source codes and data are available at: https://github.com/MLBreeding/SABO-ILSTSVR.

Bibliometric analysis of genomic selection in breeding of animal from 1993 to 2024: global trends and advancements.

Animal breeding became a difficult science when numerous genes influenced economically significant features. The major source of genetic improvement is selection, and as such, the large generation intervals in these strategies lead to reduced rates of improvement. Therefore, breeding control, genetic improvement research, and selection processes are accelerated by genomic selection. This article regarding global research interest trends in genomic selection in animal breeding themes was examined using bibliometric analysis, which employed papers from 1993 to 2024 from the SCI-Expanded, SSCI, AHCI, and E-SCI indexes. Over the period of 31years, the first 3,181 published articles on genomic selection in animal breeding were gathered. Additionally, the study displays trends in co-authorships according to nations and academic institutions as well as co-occurrences of author keywords. There have been more articles since 2010 about the use of genomic selection in animal breeding, building up a sizable library of work that will last until 2024. Among the top academics in the field are Calus MPL, Li J, and Wang Y. The most productive institutions were The United Kingdom's University of Edinburgh, Aarhus University (Denmark) and China Agricultural University. The current hotspots in this field of study include "selection," and "association," according to keyword co-occurrence and frequency analysis. China, the United States, Brazil, Canada, and United Kingdom are the top five countries that produced the most papers with the highest levels of international collaboration and networking. The main topics of current study include prediction, accuracy, association, traits, and selection. New techniques for selection, prediction, accuracy, traits, and association were developed as the discipline matured. Research collaborations across countries, institutions, and writers promote knowledge sharing, effective issue resolution, and superior outcomes.

Genetic impact of external Targhee sires at the U.S. Sheep Experiment Station: a case study of introgression.

Sheep breeders requested that the U.S. Sheep Experiment Station (USSES) to participate in national genetic evaluation through the National Sheep Improvement Program (NSIP). The reasons included the need for (1) a comparison of the productivity of industry and United States Department of Agriculture (USDA) lines, (2) transparency of USDA flocks, (3) genetic ties for NSIP by sampling of industry flocks, and (4) development of premium genetic lines for public release. In response, USSES began to incorporate external sires from NSIP participating flocks into the USSES Targhee flock. Our objective, based on a pedigree analysis, was to test if introgression of external genetics into the flock was achieved. The pedigree included 13,189 animals with mean maximum generations, mean complete generations, and mean equivalent complete generations of 4.2, 1.8, and 2.6, respectively. The mean generation interval was 3.1 yr. The reference population was defined as lambs born from 2021 to 2023 (n = 792). Two additional populations were defined as the current mature ewe flock (n = 123) and the current mature rams (n = 14). The Genetic Conservation Index averaged 7.7 for the full population and 25.7 for the reference population. Overall inbreeding was 0.003 for the full population and 0.006 for the reference population. The rate of inbreeding was 0.0003 per generation. Average relatedness was 0.015 for the full population and 0.018 for the reference population. The effective number of founders, effective number of ancestors, and founder genome equivalents contributing to the reference population were 60, 39, and 19.1, respectively. The ratio of the effective number of founders to the effective number of ancestors was 1.5, indicating the presence of genetic bottlenecks. Measures of effective population size ranged from 102 to 547. Of the 704 offspring produced by external sires, 17 ram lambs and 132 ewe lambs were retained for breeding. The USSES sires produced 299 offspring with 2 ram lambs and 51 ewe lambs retained. Incorporating external sires resulted in a cumulative percentage of genetic variance of 48.8, 49.1, and 44.2 of external genetics for the reference population, current mature ewe flock, and current mature rams, respectively. Stakeholder needs were addressed by introgression of external sires and participation in NSIP, but future selection practices need to be modified to maintain a minimum of 50% USSES core genetics in the flock.

Oocyte developmental potential and embryo production before puberty in cattle.

With the development of in vitro technologies, embryos can be produced using oocytes retrieved directly from the ovaries, i.e., regardless of ovulation. This has allowed the use of different animal categories as oocyte donors, including prepubertal cattle. The advantages of using this strategy to shorten the generation interval and accelerate genetic gain over time were soon recognized, and the first offspring generated using oocytes collected from calves were born in the early 1990s. Nevertheless, embryo production from calves and prepubertal heifers remains a challenge. The oocytes collected before puberty present low in vitro developmental potential, and the subsequent blastocyst rates are consistently lower than those from pubertal females. The acquisition of developmental competence by the oocytes occurs progressively throughout the prepubertal period, which can be subdivided into an early, intermediate, and late prepubertal (or peripubertal) phases, each characterized by different physiological and endocrine features. Therefore, embryo yield increases with age but will only achieve its maximum after puberty. The most common strategy to improve oocyte developmental potential before puberty is the use of gonadotrophic stimulation prior to oocyte retrieval. The results with superstimulation, however, vary among studies, depending on the source, dose, and length of FSH treatment, as well as the age and breed of the donors. The use of calves and prepubertal heifers as oocyte donors should also consider the possible impacts of the oocyte retrieval technique (LOPU or OPU) and the use of exogenous hormones on their subsequent fertility and productive life.

Recent advancements to increase success in assisted reproductive technologies in cattle.

Assisted reproductive technologies (ART) are fundamental for cattle breeding and sustainable food production. Together with genomic selection, these technologies contribute to reducing the generation interval and accelerating genetic progress. In this paper, we discuss advancements in technologies used in the fertility evaluation of breeding animals, and the collection, processing, and preservation of the gametes. It is of utmost importance for the breeding industry to select dams and sires of the next generation as young as possible, as is the efficient and timely collection of gametes. There is a need for reliable and easily applicable methods to evaluate sexual maturity and fertility. Although gametes processing and preservation have been improved in recent decades, challenges are still encountered. The targeted use of sexed semen and beef semen has obliterated the production of surplus replacement heifers and bull calves from dairy breeds, markedly improving animal welfare and ethical considerations in production practices. Parallel with new technologies, many well-established technologies remain relevant, although with evolving applications. In vitro production (IVP) has become the predominant method of embryo production. Although fundamental improvements in IVP procedures have been established, the quality of IVP embryos remains inferior to their in vivo counterparts. Improvements to facilitate oocyte maturation and development of new culture systems, e.g. microfluidics, are presented in this paper. New non-invasive and objective tools are needed to select embryos for transfer. Cryopreservation of semen and embryos plays a pivotal role in the distribution of genetics, and we discuss the challenges and opportunities in this field. Finally, machine learning (ML) is gaining ground in agriculture and ART. This paper delves into the utilization of emerging technologies in ART, along with the current status, key challenges, and future prospects of ML in both research and practical applications within ART.

GENOMICS AS A TOOL FOR IMPROVING DAIRY CATTLE POPULATIONS

The dairy industry plays a pivotal role in meeting the global demand for high-quality milk and dairy products. To address the ever-increasing need for efficient and productive dairy cattle populations, the integration of genomics has emerged as a transformative tool. Genomic selection, a revolutionary advancement in cattle breeding, has allowed for the identification of superior individuals at an early age based on their genetic potential. Genomic selection has not only accelerated genetic progress but has also minimized generation intervals, leading to more efficient breeding programs. Furthermore, the incorporation of genomic data in selection decisions has increased the accuracy of breeding values, ultimately resulting in the production of dairy cattle with improved milk yield, fertility, and disease resistance. Beyond traditional breeding goals, genomics has also enabled the identification of genetic variants associated with traits that are difficult to measure, such as feed efficiency, heat tolerance, and environmental adaptability. By deciphering the genetic basis of these traits, it is now possible to develop breeding strategies that enhance the resilience of dairy cattle populations in the face of changing environmental conditions. This paper also explores the role of genomics in addressing health and welfare concerns within the dairy industry. Genetic markers associated with susceptibility to diseases and stress-related conditions have been identified, providing opportunities to select for healthier and more resilient animals. Furthermore, genomics has opened avenues for tailoring nutritional strategies to individual cattle based on their genetic predispositions, optimizing feed efficiency and reducing environmental impacts. The utilization of genomic data is not limited to breeding and production. Genomics has facilitated the development of innovative tools for managing dairy cattle populations. Precision management, including genomic-based early disease detection and personalized healthcare, is becoming increasingly viable, enhancing animal welfare and reducing production losses. In conclusion, genomics has revolutionized dairy cattle breeding and management by providing accurate and efficient tools for selecting animals with superior genetic potential, improving production, health, and welfare, and enhancing the overall sustainability of dairy cattle populations. As genomics continues to evolve, the dairy industry stands to benefit from a more precise and sustainable approach to cattle production, ensuring a resilient and productive future for the dairy sector.

An FCM based weather type classification method considering photovoltaic output and meteorological characteristics and its application in power interval forecasting

AbstractWith the increasing installation of photovoltaic (PV) systems, the impact of their randomness and volatility on power system has become a significant concern. To effectively quantify the uncertainty of PV output, it is crucial to develop reliable PV power interval forecasting technology. However, the complex relationship between PV output and meteorological factors makes it challenging for a single forecasting model to meet the forecasting demand. To solve this problem, this paper proposes a weather classification method that takes into account both PV output and meteorological characteristics. Initially, the relationship between PV output and meteorological factors is analyzed, and weather types are classified using fuzzy c‐means algorithm (FCM). Then, an extreme learning machine (ELM) is employed to establish point forecasting model. By combining kernel density estimation, a PV power generation interval forecasting model is derived. The results demonstrate that the FCM‐ELM model achieves higher forecasting accuracy and narrower interval width compared to traditional ELM models, with accuracy improvement of more than 2%. Additionally, the proposed method outperforms seasonal models with an accuracy improvement of more than 1%. The contribution of this paper includes identifying the limitations of traditional weather classification methods, proposing a novel multi‐model approach for PV interval forecasting.

Pedigree diversity and implications for genetic selection of Katahdin sheep.

The Katahdin hair breed gained popularity in the United States as low input and prolific, with a propensity to exhibit parasite resistance. With the introduction of genomically enhanced estimated breeding values (GEBV) to the Katahdin genetic evaluation, defining the diversity present in the breed is pertinent. Utilizing pedigree records (n = 92,030) from 1984 to 2019 from the National Sheep Improvement Program, our objectives were to (i) estimate the completeness and quality of the pedigree, (ii) calculate diversity statistics for the whole pedigree and relevant reference subpopulations and (iii) assess the impact of current diversity on genomic selection. Reference 1 was Katahdins born from 2017 to 2019 (n = 23,494), while reference 2 was a subset with at least three generations of Katahdin ancestry (n = 9327). The completeness of the whole pedigree, and the pedigrees of reference 1 and reference 2, were above 50% through the fourth, fifth and seventh generation of ancestors, respectively. Effective population size (Ne) averaged 111 animals with a range from 42.2 to 451.0. The average generation interval was 2.9 years for the whole pedigree and reference 1, and 2.8 years for reference 2. The mean individual inbreeding and average relatedness coefficients were 1.62% and 0.91%, 1.74% and 0.90% and 2.94% and 1.46% for the whole pedigree, reference 1, and reference 2, respectively. There were over 300 effective founders in the whole pedigree and reference 1, with 169 in reference 2. Effective number of ancestors were over 150 for the whole pedigree and reference 1, while there were 67 for reference 2. Prediction accuracies increased as the reference population grew from 1k to 7.5k and plateaued at 15k animals. Given the large number of founders and ancestors contributing to the base genetic variation in the breed, the Ne is sufficient to maintain diversity while achieving progress with selection. Stable low rates of inbreeding and relatedness suggest that incorporating genetic conservation in breeding decisions is currently not of high priority. Current Ne suggests that with limited genotyping, high levels of accuracy for genomic prediction can be achieved. However, intense selection on GEBV may cause loss of genetic diversity long term.

The present of Italian Mediterranean buffalo: precision breeding based on multi-omics data

Genetic evaluation in the Italian Mediterranean Buffalo (IMB) traditionally relied on the BLUP method (best linear unbiased predictor), a mixed model system incorporating both random and fixed effects simultaneously. However, recent advancements in genome sequencing technologies have opened up the opportunity to incorporate genomic information into genetic evaluations. The ssGBLUP (single-step best linear unbiased predictor) has become the method par excellence. It replaces the traditional relationship matrix with one that combines pedigree and genomic relationships, allowing for the estimation of genetic values for non-genotyped animals. The findings of this study highlight how genomic selection enhances the precision of breeding values, facilitates greater genetic advancement and reduces the generation interval, ultimately enabling a rapid return on investment.

A transformer-based genomic prediction method fused with knowledge-guided module.

Genomic prediction (GP) uses single nucleotide polymorphisms (SNPs) to establish associations between markers and phenotypes. Selection of early individuals by genomic estimated breeding value shortens the generation interval and speeds up the breeding process. Recently, methods based on deep learning (DL) have gained great attention in the field of GP. In this study, we explore the application of Transformer-based structures to GP and develop a novel deep-learning model named GPformer. GPformer obtains a global view by gleaning beneficial information from all relevant SNPs regardless of the physical distance between SNPs. Comprehensive experimental results on five different crop datasets show that GPformer outperforms ridge regression-based linear unbiased prediction (RR-BLUP), support vector regression (SVR), light gradient boosting machine (LightGBM) and deep neural network genomic prediction (DNNGP) in terms of mean absolute error, Pearson's correlation coefficient and the proposed metric consistent index. Furthermore, we introduce a knowledge-guided module (KGM) to extract genome-wide association studies-based information, which is fused into GPformer as prior knowledge. KGM is very flexible and can be plugged into any DL network. Ablation studies of KGM on three datasets illustrate the efficiency of KGM adequately. Moreover, GPformer is robust and stable to hyperparameters and can generalize to each phenotype of every dataset, which is suitable for practical application scenarios.

Genetic Variability and Conservation Challenges in Lithuanian Dairy Cattle Populations.

The purpose of the study was to investigate the genetic variability of open Lithuanian Red and Red-and-White (LRWP) and Lithuanian Black-and-White (LBWP) dairy cattle populations and indicate the differences from the old genotypes of Lithuanian Black-and-White (LBW) and Lithuanian Red cattle (LR), which are currently under a conservation program. In order to gain a better understanding of the populations under conservation and to minimize the potential influence of other breeds, a distinct subgroup was formed that comprised animals whose father and mother belonged to the same breed (LR_pure and LBW_pure). The genetic variability was estimated using the number of founders, pedigree completeness, number of males and females in reproduction and age distribution, generation interval (GI), inbreeding coefficient (F) and effective population size (Ne). The highest average pedigree completeness values in the second generations of the old genotype LR and LBW were 100%. Higher ages of females in the populations under conservation were related to a higher GI and their longer life expectancy. In 2021, the reproductive age of bulls used for insemination within these populations ranged from 5.1 to 27.8 years. The proportions of males producing offspring in their older age indicate that the semen was used from the national gene bank of commercial artificial insemination companies. The GI (>5) in LR and LBW females was higher than that in LRWP and LBWP. The analysis of the data over the 15-year period showed that the GI of males in LRWP and LBWP decreased equally by 38%, while in LR_pure population, it increased by 80%. A high (9.24%) average inbreeding coefficient (F) was found in inbred animals of LR_pure population, while in LBW_pure, it was 5.35% in 2021. The coefficient of inbreeding varied within the different cattle populations. In the open LR population, it ranged from 1.48% to 2.7%, while in the LRWP population, it fell between 2.12% and 3.72%. The lowest effective population size (Ne) concerning the rate of inbreeding was observed in LBW_pure (23) and LR_pure (59), with the highest Ne identified in the LBWP population (462). When considering Ne based on the number of parents, LR_pure displayed the lowest Ne (42), while the highest Ne was found in LBWP (4449). An analysis of local cattle populations reveals that LR faces the most critical situation. This particular population has been steadily declining for a number of years, necessitating additional measures and efforts to safeguard the LR's ancestral genetic makeup. The results of the LBWP analysis also highlight a concerning trend. Even in very large populations with open breeding programs, the effective population size per generation can experience a significant decrease.

117 Combining Large Broiler Populations into a Single Genomic Evaluation: Dealing with Genetic Divergence

Abstract Combining populations that diverged at some point is a common practice, especially in poultry, where generation intervals are short and genetic evaluations should be available often. This study assessed whether combining large, divergent broiler populations into one genomic evaluation is feasible and the best strategy for doing that in single-step GBLUP. The pedigree for populations 1 and 2 included 397,954 and 411,780 animals, of which 154,911 and 168,429 were genotyped. Phenotypes for body weight (g), carcass yield (%), mortality (0-1), and feet health (0-6) were available for 397,954 animals in population 1 and 411,780 in population 2. A four-trait model was used in the analyses, and the genetic differences between the two populations were modeled in different ways: M1) included an extra fixed effect that accounted for population origin; M2) added origin-specific fixed effects, where the animal origin was concatenated with all fixed effects in the model. The two models were compared against a conventional model without accounting for the animal origin as a fixed effect (M0). Unknown parent groups (UPG) were fit to account for the genetic differences in M0, M1, and M2; therefore, they were set based origin population of the animal and sex. The models were evaluated based on accuracy, bias, and dispersion from the linear regression (LR) method and predictive ability (PA), where GEBV in the reduced data was compared against phenotypes adjusted for fixed effects. Validations were done separately in single and multi-populations to better assess the benefits of combining the two populations. Overall, the accuracy obtained by PA was less than by LR. No apparent accuracy, PA, bias, and dispersion differences were observed among the multi-population models across all traits. This was also true when comparing validation parameters in a multi-population against the ones from a single population. However, we observed that animals from one population were always superior to the other. A multi-trait model was applied, treating the same trait in each population differently. Genetic correlations between the same trait in different populations were moderate to low, suggesting that the two lines should possibly be evaluated separately. Defining whether populations should be combined in a single evaluation may go beyond validation parameters. Identifying if traits are the same across populations is crucial for the feasibility of combined evaluations.