Analysis of aroma precursors in Jinmudan fresh tea leaves and dynamic change of fatty acid volatile during black tea processing.

Aroma is an important factor affecting the quality of tea. Fatty acids are one of precursors and their derived contributes to tea aroma considerably. In this study, we analyzed the fatty acids of Jinmudan fresh tea leaves in different stalk position. It was found that with shoot maturity increased, the content of PUFAs (Polyunsaturated fatty acids) was increased while the content of SFAs (Saturated fatty acids) and MUFAs (Monounsaturated fatty acids) gradually decreased. During the processing period, totally 704 kinds of compounds were identified, among them, 27 kinds of fatty acid-derived volatile compounds were selected including 6 kinds of aldehydes, 8 kinds of alcohols, 13 kinds of esters and their dynamic change were revealed. Finally, the character of aroma during main processing stages and processed tea was concluded by using a flavor wheel. This study results provide a theoretical basis for the improvement of processing and quality in Jinmudan black tea.

Genome-wide identification of CsCCD gene family in tea plant (Camellia sinensis) and expression analysis of the oolong tea processing with supplementary LED light / 生物工程学报

Carotenoid cleavage dioxygenase (CCD) family is important for production of volatile aromatic compounds and synthesis of plant hormones. To explore the biological functions and gene expression patterns of CsCCD gene family in tea plant, genome-wide identification of CsCCD gene family was performed. The gene structures, conserved motifs, chromosome locations, protein physicochemical properties, evolutionary characteristics, interaction network and cis-acting regulatory elements were predicted and analyzed. Real time-quantitative reverse transcription PCR (RT-qPCR) was used to detect the relative expression level of CsCCD gene family members under different leaf positions and light treatments during processing. A total of 11 CsCCD gene family members, each containing exons ranging from 1 to 11 and introns ranging from 0 to 10, were identified. The average number of amino acids and molecular weight were 519 aa and 57 643.35 Da, respectively. Phylogenetic analysis showed the CsCCD gene family was clustered into 5 major groups (CCD1, CCD4, CCD7, CCD8 and NCED). The CsCCD gene family mainly contained stress response elements, hormone response elements, light response elements and multi-factor response elements, and light response elements was the most abundant (142 elements). Expression analysis showed that the expression levels of CsCCD1 and CsCCD4 in elder leaves were higher than those in younger leaves and stems. With the increase of turning over times, the expression levels of CsCCD1 and CsCCD4 decreased, while supplementary LED light strongly promoted their expression levels in the early stage. The expression level of NCED in younger leaves was higher than that in elder leaves and stems on average, and the expression trend varied in the process of turning over. NCED3 first increased and then decreased, with an expression level 15 times higher than that in fresh leaves. In the late stage of turning over, supplementary LED light significantly promoted its gene expression. In conclusion, CsCCD gene family member expressions were regulated by mechanical force and light. These understandings may help to optimize tea processing techniques and improve tea quality.

On the Generation of Medical Dialogues for COVID-19

Under the pandemic of COVID-19, people experiencing COVID19-related symptoms or exposed to risk factors have a pressing need to consult doctors. Due to hospital closure, a lot of consulting services have been moved online. Because of the shortage of medical professionals, many people cannot receive online consultations timely. To address this problem, we aim to develop a medical dialogue system that can provide COVID19-related consultations. We collected two dialogue datasets - CovidDialog - (in English and Chinese respectively) containing conversations between doctors and patients about COVID-19. On these two datasets, we train several dialogue generation models based on Transformer, GPT, and BERT-GPT. Since the two COVID-19 dialogue datasets are small in size, which bear high risk of overfitting, we leverage transfer learning to mitigate data deficiency. Specifically, we take the pretrained models of Transformer, GPT, and BERT-GPT on dialog datasets and other large-scale texts, then finetune them on our CovidDialog datasets. Experiments demonstrate that these approaches are promising in generating meaningful medical dialogues about COVID-19. But more advanced approaches are needed to build a fully useful dialogue system that can offer accurate COVID-related consultations. The data and code are available at https://github.com/UCSD-AI4H/COVID-Dialogue