ABSTRACT
The infection caused by Nosema bombycis often known as pebrine, is a devastating sericulture disease. The infection can be transmitted to the next generation through eggs laid by infected female Bombyx mori moths (transovarial) as well as with N. bombycis contaminated food (horizontal). Most diagnoses were carried out in the advanced stages of infection until the time that infection might spread to other healthy insects. Hence, early diagnosis of pebrine is of utmost importance to quarantine infected larvae from uninfected silkworm batches and stop further spread of the infection. The findings of our study provide an insight into how the silkworm larval host defence system was activated against early N. bombycis transovarial infection. The results obtained from transcriptome analysis of infected 2nd instar larvae revealed significant (adjusted P-value < 0.05) expression of 1888 genes of which 801 genes were found to be upregulated and 1087 genes were downregulated when compared with the control. Pathway analysis indicated activation of the immune deficiency (IMD) pathway, which shows a potential immune defence response against pebrine infection as well as suppression of the melanin synthesis pathway due to lower expression of prophenoloxidase activating enzyme (PPAE). Liquid chromatography mass spectrometry (LC-MS/MS) analysis of haemolymph from infected larvae shows the secretion of serpin binding protein of N. bombycis which might be involved in the suppression of the melanization pathway. Moreover, among the differentially expressed genes, we found that LPMC-61, yellow-y, gasp and osiris 9 can be utilised as potential markers for early diagnosis of transovarial pebrine infection in B. mori. Physiological as well as biochemical roles and functions of many of the essential genes are yet to be established, and enlightened research will be required to characterize the products of these genes.
ABSTRACT
An electronic health record (EHR) is a vital high-dimensional part of medical concepts. Discovering implicit correlations in the information of this data set and the research and informative aspects can improve the treatment and management process. The challenge of concern is the data sources' limitations in finding a stable model to relate medical concepts and use these existing connections. This paper presents Patient Forest, a novel end-to-end approach for learning patient representations from tree-structured data for readmission and mortality prediction tasks. By leveraging statistical features, the proposed model is able to provide an accurate and reliable classifier for predicting readmission and mortality. Experiments on MIMIC-III and eICU datasets demonstrate Patient Forest outperforms existing machine learning models, especially when the training data are limited. Additionally, a qualitative evaluation of Patient Forest is conducted by visualising the learnt representations in 2D space using the t-SNE, which further confirms the effectiveness of the proposed model in learning EHR representations.
