RESUMEN
Background: Understanding social determinants of health (SDOH) that may be risk factors for childhood obesity is important to developing targeted interventions to prevent obesity. Prior studies have examined these risk factors, mostly examining obesity as a static outcome variable. Methods: We extracted electronic health record data from 2012 to 2019 for a children's health system that includes two hospitals and wide network of outpatient clinics spanning five East Coast states in the United States. Using data-driven and algorithmic clustering, we have identified distinct BMI-percentile classification groups in children from 0 to 7 years of age. We used two separate algorithmic clustering methods to confirm the robustness of the identified clusters. We used multinomial logistic regression to examine the associations between clusters and 27 neighborhood SDOHs and compared positive and negative SDOH characteristics separately. Results: From the cohort of 36,910 children, five BMI-percentile classification groups emerged: always having obesity (n = 429; 1.16%), overweight most of the time (n = 15,006; 40.65%), increasing BMI percentile (n = 9,060; 24.54%), decreasing BMI percentile (n = 5,058; 13.70%), and always normal weight (n = 7,357; 19.89%). Compared to children in the decreasing BMI percentile and always normal weight groups, children in the other three groups were more likely to live in neighborhoods with higher poverty, unemployment, crowded households, single-parent households, and lower preschool enrollment. Conclusions: Neighborhood-level SDOH factors have significant associations with children's BMI-percentile classification and changes in classification. This highlights the need to develop tailored obesity interventions for different groups to address the barriers faced by communities that can impact the weight and health of children living within them.
RESUMEN
Cyclooxygenase-2 (COX-2) is an enzyme that plays a pivotal role in peripheral inflammation and pain via the prostaglandin pathway. In the central nervous system (CNS), COX-2 is implicated in neurodegenerative and psychiatric disorders as a potential therapeutic target and biomarker. However, clinical studies with COX-2 have yielded inconsistent results, partly due to limited mechanistic understanding of how COX-2 activity relates to CNS pathology. Therefore, developing COX-2 positron emission tomography (PET) radiotracers for human neuroimaging is of interest. This study introduces [11C]BRD1158, which is a potent and uniquely fast-binding, selective COX-2 PET radiotracer. [11C]BRD1158 was developed by prioritizing potency at COX-2, isoform selectivity over COX-1, fast binding kinetics, and free fraction in the brain. Evaluated through in vivo PET neuroimaging in rodent models with human COX-2 overexpression, [11C]BRD1158 demonstrated high brain uptake, fast target-engagement, functional reversibility, and excellent specific binding, which is advantageous for human imaging applications. Lastly, post-mortem samples from Huntington's disease (HD) patients and preclinical HD mouse models showed that COX-2 levels were elevated specifically in disease-affected brain regions, primarily from increased expression in microglia. These findings indicate that COX-2 holds promise as a novel clinical marker of HD onset and progression, one of many potential applications of [11C]BRD1158 human PET.
RESUMEN
Background Nursing professionals, comprising the largest workforce engaged in the primary healthcare system, play a pivotal role in addressing population health needs. However, gaps in the training of nurses and midwives in lower-middle-income countries may undermine their performance and necessary skill development for fulfilling key population health needs. Substantial challenges exist in improving the regular curricular and refresher training of diplomate nurses and midwives working in primary care facilities and supporting both clinical care and health promotion functions. The study objective was to conduct a gap analysis in the present nursing curriculum and training profile of general duty midwives working in urban primary health facilities and understand their expectations and preferences from the planned refresher training course. Methods We conducted a qualitative explorative study among General Nursing midwives (GNMs) working in urban primary health facilities in the Gurugram district of Haryana, India to conduct a gap analysis in their present curriculum and training preferences. Results A total of 17 nurses with a mean (SD) age of 33.52 (4.75) years and an average nursing work experience of 5.35 (0.56) years were interviewed in-depth. Lack of practical applicability, complex study material, inexperienced tutors, and weak English language comprehension were key barriers in the existing nursing curriculum. The nurses expressed willingness to participate in refresher training with varied expectations, although there existed a distinct preference for short, flexible, and blended online-offline modes of training. Conclusions Strengthening GNM nursing education should be prioritized in Indian health settings with the focus on improving student comprehension through vernacular instruction when feasible, and capacity building of tutors, with avenues for continued training and education. There is also a need for strengthening the curriculum related to key emergent public health challenges related to non-communicable diseases and mental health, as also skills for client and patient counseling and communication.
RESUMEN
The lack of resistance to Lipaphis erysimi in cultivated Brassicas makes caused this pest highly devastating resulting in significant loss of rapeseed-mustard productivity in India. B. fruticulosa, a wild crucifer is known as an excellent source of resistance to L. erysimi. Therefore, we planned to assess defense associated biochemical alterations and molecular components of B. juncea-B. fruticulosa ILs to mustard aphid. Phenotypic assessment of ILs on the basis of aphid population per plant (APP) categorized genotypes into resistant (7.15-18.50 APP), moderately susceptible (42.29-53.33 APP) and susceptible (70.00-77.07 APP) genotypes. Mustard aphid infested minimally B. fruticulosa (0.80 APP) among tested genotypes. The maximum increase in catalase (CAT) activity was determined in B. fruticulosa and resistant ILs after 48 h (2.03 and 1.76-fold, respectively) and one week (2.98 and 1.79-fold, respectively) of mustard aphid infestation. The strong induction of CAT2 transcripts (19.25-fold) and CAT activity (5.88-fold) along with low aphid count in resistant IL, Ad4-64 (13.85 APP) suggested the pivotal role of CAT in resistance to mustard aphid. Guaiacol peroxidase (GPX) was significantly decreased following pest infestation at both infestation stages. The ascorbate content was highest in resistant IL, ADV-6RD (2.14-fold) after one week of aphid infestation. H2O2 content rapidly increased in B. juncea-B. fruticulosa derived lines after 48 h of aphid infestation. The negative and significant association between APP and CAT (- 0.56** and - 0.48*, respectively), glutathione (- 0.43* and - 0.40*, respectively), H2O2 (- 0.57** and - 0.43*, respectively) at both 48 h and one week infestation stages signified their role in deterring mustard aphid infestation. The positive and significant association between total sugars (0.33* at 7 DPI), reducing sugars (0.33* at 7 DPI), sucrose (0.36** at 48 h) and APP indicated that higher the sugars content, higher will be mustard aphid infestation in B. juncea derived ILs. The information being generated and key candidates (CAT2, ascorbate and H2O2) being identified may help in effective deployment of B. fruticulosa resistance in mustard breeding.
RESUMEN
Background: Understanding social determinants of health (SDOH) that may be risk factors for childhood obesity is important to developing targeted interventions to prevent obesity. Prior studies have examined these risk factors, mostly examining obesity as a static outcome variable. Objectives: This study aimed to identify distinct subpopulations based on BMI percentile classification or changes in BMI percentile classifications over time and explore these longitudinal associations with neighborhood-level SDOH factors in children from 0 to 7 years of age. Methods: Using Latent Class Growth (Mixture) Modelling (LCGMM) we identify distinct BMI% classification groups in children from 0 to 7 years of age. We used multinomial logistic regression to study associations between SDOH factors with each BMI% classification group. Results: From the study cohort of 36,910 children, five distinct BMI% classification groups emerged: always having obesity (n=429; 1.16%), overweight most of the time (n=15,006; 40.65%), increasing BMI% (n=9,060; 24.54%), decreasing BMI% (n=5,058; 13.70%), and always normal weight (n=7,357; 19.89%). Compared to children in the decreasing BMI% and always normal weight groups, children in the other three groups were more likely to live in neighborhoods with higher rates of poverty, unemployment, crowded households, and single-parent households, and lower rates of preschool enrollment. Conclusions: Neighborhood-level SDOH factors have significant associations with children's BMI% classification and changes in classification over time. This highlights the need to develop tailored obesity interventions for different groups to address the barriers faced by communities that can impact the weight and health of the children living within them.
RESUMEN
Thyrotoxic hypokalaemic periodic paralysis (THPP) is a rare complication of hyperthyroidism that is potentially life-threatening if not treated promptly. It is more common in Asian and Polynesian populations and very few cases have been reported to date in people of White ethnicity. We present a case report of a young male patient of White ethnicity, who was initially brought in as a stroke alert with tetraparesis which was ruled out on initial assessment, but then had a syncopal episode and was noted to be initially bradycardic and subsequently tachycardic. Blood tests showed hypokalaemia and hypophosphataemia and he was treated as a hypokalaemic periodic paralysis patient. Intravenous potassium replacement was commenced. Symptoms and ECG changes resolved with correction of potassium levels. Thyroid function tests requested later were suggestive of hyperthyroidism and the diagnosis of thyrotoxic hypokalaemic periodic paralysis was made. This is an interesting case given its rarity, and this case report highlights the importance of early diagnosis and prompt treatment.
RESUMEN
Various types of machine learning techniques are available for analyzing electronic health records (EHRs). For predictive tasks, most existing methods either explicitly or implicitly divide these time-series datasets into predetermined observation and prediction windows. Patients have different lengths of medical history and the desired predictions (for purposes such as diagnosis or treatment) are required at different times in the future. In this paper, we propose a method that uses a sequence-to-sequence generator model to transfer an input sequence of EHR data to a sequence of user-defined target labels, providing the end-users with "flexible" observation and prediction windows to define. We use adversarial and semi-supervised approaches in our design, where the sequence-to-sequence model acts as a generator and a discriminator distinguishes between the actual (observed) and generated labels. We evaluate our models through an extensive series of experiments using two large EHR datasets from adult and pediatric populations. In an obesity predicting case study, we show that our model can achieve superior results in flexible-window prediction tasks, after being trained once and even with large missing rates on the input EHR data. Moreover, using a number of attention analysis experiments, we show that the proposed model can effectively learn more relevant features in different prediction tasks.
RESUMEN
Holt-Oram syndrome is a rare autosomal dominant disorder which occurs because of mutations in the TBX5 genes. Most notable manifestations include musculoskeletal deformities, predominantly affecting the upper limbs, and congenital heart defects. Presentation could be multifaceted leading to delay in diagnosis. We describe an interesting incidental diagnosis of Holt-Oram syndrome in a young female adult who accompanied her son to the clinic. He had undergone closure of both atrial septal defect (ASD) and patent ductus arteriosus (PDA) in his infancy. She reported progressive exertional dyspnoea, reduced exercise tolerance, and palpitations; incidentally, she was noted to have right upper limb deformities. These findings prompted further evaluation and thereafter, resulted in a diagnosis of Holt-Oram syndrome.
RESUMEN
Childhood obesity is a major public health challenge. Early prediction and identification of the children at an elevated risk of developing childhood obesity may help in engaging earlier and more effective interventions to prevent and manage obesity. Most existing predictive tools for childhood obesity primarily rely on traditional regression-type methods using only a few hand-picked features and without exploiting longitudinal patterns of children's data. Deep learning methods allow the use of high-dimensional longitudinal datasets. In this paper, we present a deep learning model designed for predicting future obesity patterns from generally available items on children's medical history. To do this, we use a large unaugmented electronic health records dataset from a large pediatric health system in the US. We adopt a general LSTM network architecture and train our proposed model using both static and dynamic EHR data. To add interpretability, we have additionally included an attention layer to calculate the attention scores for the timestamps and rank features of each timestamp. Our model is used to predict obesity for ages between 3-20 years using the data from 1-3 years in advance. We compare the performance of our LSTM model with a series of existing studies in the literature and show it outperforms their performance in most age ranges.
RESUMEN
An increasing amount of research is being devoted to applying machine learning methods to electronic health record (EHR) data for various clinical purposes. This growing area of research has exposed the challenges of the accessibility of EHRs. MIMIC is a popular, public, and free EHR dataset in a raw format that has been used in numerous studies. The absence of standardized preprocessing steps can be, however, a significant barrier to the wider adoption of this rare resource. Additionally, this absence can reduce the reproducibility of the developed tools and limit the ability to compare the results among similar studies. In this work, we provide a greatly customizable pipeline to extract, clean, and preprocess the data available in the fourth version of the MIMIC dataset (MIMIC-IV). The pipeline also presents an end-to-end wizard-like package supporting predictive model creations and evaluations. The pipeline covers a range of clinical prediction tasks which can be broadly classified into four categories - readmission, length of stay, mortality, and phenotype prediction. The tool is publicly available at https://github.com/healthylaife/MIMIC-IV-Data-Pipeline.
RESUMEN
With the growing availability of Electronic Health Records (EHRs), many deep learning methods have been developed to leverage such datasets in medical prediction tasks. Notably, transformer-based architectures have proven to be highly effective for EHRs. Transformer-based architectures are generally very effective in "transferring" the acquired knowledge from very large datasets to smaller target datasets through their comprehensive "pre-training" process. However, to work efficiently, they still rely on the target datasets for the downstream tasks, and if the target dataset is (very) small, the performance of downstream models can degrade rapidly. In biomedical applications, it is common to only have access to small datasets, for instance, when studying rare diseases, invasive procedures, or using restrictive cohort selection processes. In this study, we present CEHR-GAN-BERT, a semi-supervised transformer-based architecture that leverages both in- and out-of-cohort patients to learn better patient representations in the context of few-shot learning. The proposed method opens new learning opportunities where only a few hundred samples are available. We extensively evaluate our method on four prediction tasks and three public datasets showing the ability of our model to achieve improvements upwards of 5% on all performance metrics (including AUROC and F1 Score) on the tasks that use less than 200 annotated patients during the training process.
RESUMEN
Acid phosphatases (Apases) are an important group of enzymes that hydrolyze soil and plant phosphoesters and anhydrides to release Pi (inorganic phosphate) for plant acquisition. Their activity is strongly correlated to the phosphorus use efficiency (PUE) of plants. Indian mustard (Brassica juncea L. Czern & Coss) is a major oilseed crop that also provides protein for the animal feed industry. It exhibits low PUE. Understanding the genetics of PUE and its component traits, especially Apase activity, will help to reduce Pi fertilizer application in the crop. In the present study, we evaluated 280 genotypes of the diversity fixed foundation set of Indian mustard for Apase activity in the root (RApase) and leaf (LApase) tissues at three- low (5µM), normal (250µM) and high (1mM) Pi levels in a hydroponic system. Substantial effects of genotype and Pi level were observed for Apase activity in both tissues of the evaluated lines. Low Pi stress induced higher mean RApase and LApase activities. However, mean LApase activity was relatively more than mean RApase at all three Pi levels. JM06016, IM70 and Kranti were identified as promising genotypes with higher LApase activity and increased R/S at low Pi. Genome-wide association study revealed 10 and 4 genomic regions associated with RApase and LApase, respectively. Annotation of genomic regions in the vicinity of peak associated SNPs allowed prediction of 15 candidates, including genes encoding different family members of the acid phosphatase such as PAP10 (purple acid phosphatase 10), PAP16, PNP (polynucleotide phosphorylase) and AT5G51260 (HAD superfamily gene, subfamily IIIB acid phosphatase) genes. Our studies provide an understanding of molecular mechanism of the Apase response of B. juncea at varying Pi levels. The identified SNPs and candidate genes will support marker-assisted breeding program for improving PUE in Indian mustard. This will redeem the crop with enhanced productivity under restricted Pi reserves and degrading agro-environments.
RESUMEN
Ectopic prostate tissue is a rare phenomenon. Histologically and immunohistochemically it is indistinguishable from normal prostatic tissue but it is difficult to recognize and may be confused with malignancy when presenting as ectopic lesion in the urinary system. Therefore, awareness of this entity is essential for both urologists and pathologists to prevent misdiagnosis.
Asunto(s)
Coristoma , Neoplasias de la Vejiga Urinaria , Coristoma/diagnóstico , Coristoma/patología , Humanos , Conducta Imitativa , Masculino , Pelvis , Próstata/patología , Próstata/cirugía , Neoplasias de la Vejiga Urinaria/diagnóstico , Neoplasias de la Vejiga Urinaria/patologíaRESUMEN
Background and Aims: Diabetic retinopathy is the most common diabetes-associated microvascular complication and is among the leading causes of vision loss or blindness in the adult population. The present study is a retrospective study that reported the natural history of diabetic retinopathy. Methods: Retrospective medical records of 170 patients aged > 20 years with a confirmed complication of diabetic retinopathy were recruited into the present study. A questionnaire was also sent to each subject for gathering their experiences, and verification was done by the attending medical physicians. The questionnaire was answered by all recruited patients. Results: The results showed that 23 (13.5%) subjects have a family history of diabetic retinopathy with 10 (5.9%) having mild NPDR, 63 (37.1%) with moderate NPDR, 60 (25.3%) have severe NPDR while 37 (21.8%) have PDR complications. The presence of co-morbidities was found in 139 (81.8%) subjects. Patients with PDR reported a significantly longer duration of diabetes mellitus with worse glycemic control. Conclusions: The study revealed and concluded that adherence to the prescribed management regimen is important, for which patient education was the key which was lacking.
Asunto(s)
Diabetes Mellitus Tipo 2 , Retinopatía Diabética , Adulto , Diabetes Mellitus Tipo 2/complicaciones , Diabetes Mellitus Tipo 2/epidemiología , Retinopatía Diabética/epidemiología , Retinopatía Diabética/etiología , Humanos , Estudios Retrospectivos , Encuestas y CuestionariosRESUMEN
Working with electronic health records (EHRs) is known to be challenging due to several reasons. These reasons include not having: 1) similar lengths (per visit), 2) the same number of observations (per patient), and 3) complete entries in the available records. These issues hinder the performance of the predictive models created using EHRs. In this paper, we approach these issues by presenting a model for the combined task of imputing and predicting values for the irregularly observed and varying length EHR data with missing entries. Our proposed model (dubbed as Bi-GAN) uses a bidirectional recurrent network in a generative adversarial setting. In this architecture, the generator is a bidirectional recurrent network that receives the EHR data and imputes the existing missing values. The discriminator attempts to discriminate between the actual and the imputed values generated by the generator. Using the input data in its entirety, Bi-GAN learns how to impute missing elements in-between (imputation) or outside of the input time steps (prediction). Our method has three advantages to the state-of-the-art methods in the field: (a) one single model performs both the imputation and prediction tasks; (b) the model can perform predictions using time-series of varying length with missing data; (c) it does not require to know the observation and prediction time window during training and can be used for the predictions with different observation and prediction window lengths, for short- and long-term predictions. We evaluate our model on two large EHR datasets to impute and predict body mass index (BMI) values and show its superior performance in both settings.
RESUMEN
Aberrant activation of NLRP3 inflammasome has been implicated in several inflammatory diseases. Autophagy is one of the primary mechanisms that regulate NLRP3 inflammasome activity. In this study, we attempted to target NLRP3 inflammasome activity by a synthetic compound IIIM-941. We found that IIIM-941 inhibits ATP induced NLRP3 inflammasome by induction of autophagy through AMPK pathway in bone marrow derived macrophages (BMDMs) and J774A.1 cells. It was interesting to observe that IIIM-941 did not show any inhibitory activity against LPS induced pro-inflammatory cytokines TNF-α and IL-6. The anti-NLRP3 activity of IIIM-941 was significantly reversed when we attempted to block autophagy by using either pharmacological inhibitor bafilomycin A1or by using siRNA against AMPK. Further, we found that IIIM-941 downregulated the expression of NLRP3 and prevented the oligomerization of ASC to exert its anti-NLRP3 inflammasome effect in J774A.1 cells. We validated inhibitory activity of IIIM-941 against NLRP3 in three different mice models. The anti-inflammatory effect of IIIM-941 was highly significant in ATP induced peritoneal inflammation model. IIIM-941 was similarly effective in suppressing MSU induced IL-1ß in the air pouch model of inflammation without affecting the levels of TNF-α and IL-6. Finally, oral efficacy of IIIM-941 was also proved in MSU indued foot paw edema model of inflammation in mice at 10 and 20 mg/kg (b.w.). The compounds like IIIM-941 can be explored further for the development of therapies against diseases such as Alzheimer's disease and Parkinson's disease, where hampered autophagy and NLRP3 activation play a crucial role in the pathological development.
RESUMEN
KEY MESSAGE: Heavy doses of gamma irradiation can reduce linkage drag by disrupting large sized alien translocations and promoting exchanges between crop and wild genomes. Resistance to mustard aphid (Lipaphis erysimi) infestation was significantly improved in Brassica juncea through B. juncea-B. fruticulosa introgression. However, linkage drag caused by introgressed chromatin fragments has so far prevented the deployment of this resistance source in commercial cultivars. We investigated the patterns of donor chromatin segment substitutions in the introgression lines (ILs) through genomic in situ hybridization (GISH) coupled with B. juncea chromosome-specific oligonucleotide probes. These allowed identification of large chromosome translocations from B. fruticulosa in the terminal regions of chromosomes A05, B02, B03 and B04 in three founder ILs (AD-64, 101 and 104). Only AD-101 carried an additional translocation at the sub-terminal to intercalary position in both homologues of chromosome A01. We validated these translocations with a reciprocal blast hit analysis using shotgun sequencing of three ILs and species-specific contigs/scaffolds (kb sized) from a de novo assembly of B. fruticulosa. Alien segment substitution on chromosome A05 could not be validated. Current studies also endeavoured to break linkage drag by exposing seeds to a heavy dose (200kR) of gamma radiation. Reduction in the size of introgressed chromatin fragments was observed in many M3 plants. There was a complete loss of the alien chromosome fragment in one instance. A few M3 plants with novel patterns of chromosome segment substitutions displayed improved agronomic performance coupled with resistance to mustard aphid. SNPs in such genomic spaces should aid the development of markers to track introgressed DNA and allow application in plant breeding.
Asunto(s)
Áfidos/fisiología , Mapeo Cromosómico/métodos , Cromosomas de las Plantas/genética , Resistencia a la Enfermedad/inmunología , Planta de la Mostaza/genética , Enfermedades de las Plantas/inmunología , Proteínas de Plantas/metabolismo , Animales , Resistencia a la Enfermedad/genética , Resistencia a la Enfermedad/efectos de la radiación , Rayos gamma , Regulación de la Expresión Génica de las Plantas , Planta de la Mostaza/crecimiento & desarrollo , Planta de la Mostaza/parasitología , Planta de la Mostaza/efectos de la radiación , Enfermedades de las Plantas/genética , Enfermedades de las Plantas/parasitología , Proteínas de Plantas/genéticaRESUMEN
NLRP3 inflammasome is involved in several chronic inflammatory diseases. The inflammatory effect of the NLRP3 inflammasome is executed through IL-1ß and IL-18. Therefore, IL-1ß is one of the primary targets in chronic inflammatory conditions. However, current treatment regimens are dependent on anti- IL-1ß biologicals. The therapies targeting IL-1ß through inhibition of NLRP3 inflammasome are thus being actively explored. We identified safranal, a small molecule responsible for the essence of saffron as a potential inhibitor of the NLRP3 inflammasome. Safranal significantly suppressed the release of IL-1ß from ATP stimulated J774A.1 and bone marrow-derived macrophages (BMDMs) by regulating CASP1 and CASP8 dependent cleavage of pro-IL-1ß. Safranal markedly suppressed the expression of NLRP3 and its ATPase activity. Safranal treatment enhanced the expression of NRF2, whereas, si-RNA mediated silencing of Nrf2 abrogated the anti-NLRP3 effect of safranal. Furthermore, safranal inhibited ASC oligomerization and formation of ASC specks. Safranal also displayed anti-NLRP3 activity in multiple mice models. Treatment of animals with safranal reduced the production of IL-1ß in ATP elicited peritoneal inflammation, MSU induced air pouch inflammation, and MSU injected foot paw edema in mice. Thus, our data projects safranal as a potential preclinical drug candidate against NLRP3 inflammasome triggered chronic inflammation.
Asunto(s)
Proteínas Adaptadoras de Señalización CARD/antagonistas & inhibidores , Proteínas Adaptadoras de Señalización CARD/metabolismo , Ciclohexenos/farmacología , Proteína con Dominio Pirina 3 de la Familia NLR/antagonistas & inhibidores , Proteína con Dominio Pirina 3 de la Familia NLR/metabolismo , Terpenos/farmacología , Animales , Línea Celular , Células Cultivadas , Ciclohexenos/uso terapéutico , Relación Dosis-Respuesta a Droga , Femenino , Humanos , Inflamación/inducido químicamente , Inflamación/tratamiento farmacológico , Inflamación/metabolismo , Ratones , Ratones Endogámicos BALB C , Terpenos/uso terapéuticoRESUMEN
Machine learning algorithms have been widely used to capture the static and temporal patterns within electronic health records (EHRs). While many studies focus on the (primary) prevention of diseases, primordial prevention (preventing the factors that are known to increase the risk of a disease occurring) is still widely under-investigated. In this study, we propose a multi-target regression model leveraging transformers to learn the bidirectional representations of EHR data and predict the future values of 11 major modifiable risk factors of cardiovascular disease (CVD). Inspired by the proven results of pre-training in natural language processing studies, we apply the same principles on EHR data, dividing the training of our model into two phases: pre-training and fine-tuning. We use the fine-tuned transformer model in a "multi-target regression" theme. Following this theme, we combine the 11 disjoint prediction tasks by adding shared and target-specific layers to the model and jointly train the entire model. We evaluate the performance of our proposed method on a large publicly available EHR dataset. Through various experiments, we demonstrate that the proposed method obtains a significant improvement (12.6% MAE on average across all 11 different outputs) over the baselines.
RESUMEN
KEY MESSAGE: Genome wide association studies allowed prediction of 17 candidate genes for association with nitrogen use efficiency. Novel information obtained may provide better understanding of genomic controls underlying germplasm variations for this trait in Indian mustard. Nitrogen use efficiency (NUE) of Indian mustard (Brassica juncea (L.) Czern & Coss.) is low and most breeding efforts to combine NUE with crop performance have not succeeded. Underlying genetics also remain unexplored. We tested 92 SNP-genotyped inbred lines for yield component traits, N uptake efficiency (NUPEFF), nitrogen utilization efficiency (NUTEFF), nitrogen harvest index (NHI) and NUE for two years at two nitrogen doses (No without added N and N100 added @100 kg/ha). Genotypes IC-2489-88, M-633, MCP-632, HUJM 1080, GR-325 and DJ-65 recorded high NUE at low N. These also showed improved crop performance under high N. One determinate mustard genotype DJ-113 DT-3 revealed maximum NUTEFF. Genome wide association studies (GWAS) facilitated recognition of 17 quantitative trait loci (QTLs). Environment specificity was high. B-genome chromosomes (B02, B03, B05, B07 and B08) harbored many useful loci. We also used regional association mapping (RAM) to supplement results from GWAS. Annotation of the genomic regions around peak SNPs helped to predict several gene candidates for root architecture, N uptake, assimilation and remobilization. CAT9 (At1g05940) was consistently envisaged for both NUE and NUPEFF. Major N transporter genes, NRT1.8 and NRT3.1 were predicted for explaining variation for NUTEFF and NUPEFF, respectively. Most significant amino acid transporter gene, AAP1 appeared associated with NUE under limited N conditions. All these candidates were predicted in the regions of high linkage disequilibrium. Sequence information of the predicted candidate genes will permit development of molecular markers to aid breeding for high NUE.