Promoter prediction in nannochloropsis based on densely connected convolutional neural networks.

Promoter is a key DNA element located near the transcription start site, which regulates gene transcription by binding RNA polymerase. Thus, the identification of promoters is an important research field in synthetic biology. Nannochloropsis is an important unicellular industrial oleaginous microalgae, and at present, some studies have identified some promoters with specific functions by biological methods in Nannochloropsis, whereas few studies used computational methods. Here, we propose a method called DNPPro (DenseNet-Predict-Promoter) based on densely connected convolutional neural networks to predict the promoter of Nannochloropsis. First, we collected promoter sequences from six Nannochloropsis strains and removed 80% similarity using CD-HIT for each strain to yield a reliable set of positive datasets. Then, in order to construct a robust classifier, within-group scrambling method was used to generate negative dataset which overcomes the limitation of randomly selecting a non-promoter region from the same genome as a negative sample. Finally, we constructed a densely connected convolutional neural network, with the sequence one-hot encoding as the input. Compared with commonly used sequence processing methods, DNPPro can extract long sequence features to a greater extent. The cross-strain experiment on independent dataset verifies the generalization of our method. At the same time, T-SNE visualization analysis shows that our method can effectively distinguish promoters from non-promoters.

[Feasibility of Preheating at 41 ℃ to Correct Red Blood Cell Parameters in the Presence of High-titer Cold Agglutinins].

Objective To explore the feasibility of preheating in 41 ℃ water bath for 30 minutes to correct the red blood cell parameters in the specimens containing high-titer cold agglutinins(CAs). Methods Two specimens containing high-titer CAs were selected during work,and the parameters of complete blood count at room temperature or after preheating in 37 ℃ or 41 ℃ water bath were compared.The smears were stained,and the distribution of red blood cells was observed with a microscope.Further,74 specimens without CAs were collected for complete blood count,and then the test results at room temperature and after preheating at 41 ℃ were compared. Results At room temperature,the specimens containing high-titer CAs showed significantly reduced red blood cell count(RBC)and hematocrit(HCT),abnormally increased mean corpuscular hemoglobin(MCH)and mean cell hemoglobin concentration(MCHC),abnormal percents of hemoglobin(HGB)and RBC,and aggregation of a large number of red blood cells.After being preheated at 37 ℃ for a certain time,the specimens demonstrated obviously improved parameters while still aggregation of a small number of red blood cells.After being preheated at 41 ℃ for 30 minutes,the specimens showed significantly increased RBC,normal HCT,MCH,and MCHC,and evenly distributed red blood cells.The 74 specimens without CAs showed the comparability was ≥80% between room temperature and preheating at 41 ℃ for 30 minutes or 60 minutes. Conclusion We can preheat the specimens containing high-titer CAs in a water bath at 41 ℃ to obtain accurate red blood cell parameters.

Composition identification and UV-C irradiation growth inhibition effect of green shading on the greenhouse cover.

Greenhouse cover pollution with green shading composed of dust, microalgae and bacteria is a severe problem in tropical areas. The shading results in lower greenhouse indoor light intensity reducing the yield and quality of protected horticulture crops. However, few studies have focused on environmentally efficient ways to remove green shading to increase greenhouse production. In this study, five purified microalgae were isolated from the green shading of three greenhouse roofs and were identified using morphological and molecular assessments. The effects of Ultraviolet-C irradiation (UV-C, 254 nm) at doses of 100, 200 and 300 mJ cm-2 on the growth of GLY-1 microalgae were investigated. The results indicated that five purified microalgae all appeared to belong to the genus of Jaagichlorella. The purified microalgae cell density and chlorophyll content decreased respectively by 26.89-74.44 % and 42.02-77.31 % at 1-3 d after UV-C treatment with doses ranging from 100 to 300 mJ cm-2. The inhibition of the growth rate of microalgae was significantly positively correlated with the UV-C irradiation dose and significantly negatively correlated with treatment time. In summary, UV-C irradiation treatment at 300 mJ cm-2 and 3 d could substantially inhibit microalgae growth in green shading on greenhouse covers. UV-C irradiation could be an effective method for solving the problem of greenhouse cover pollution with microalgae.