Patch-Based Convolutional Encoder: A Deep Learning Algorithm for Spectral Classification Balancing the Local and Global Information.

Molecular vibrational spectroscopies, including infrared absorption and Raman scattering, provide molecular fingerprint information and are powerful tools for qualitative and quantitative analysis. They benefit from the recent development of deep-learning-based algorithms to improve the spectral, spatial, and temporal resolutions. Although a variety of deep-learning-based algorithms, including those to simultaneously extract the global and local spectral features, have been developed for spectral classification, the classification accuracy is still far from satisfactory when the difference becomes very subtle. Here, we developed a lightweight algorithm named patch-based convolutional encoder (PACE), which effectively improved the accuracy of spectral classification by extracting spectral features while balancing local and global information. The local information was captured well by segmenting the spectrum into patches with an appropriate patch size. The global information was extracted by constructing the correlation between different patches with depthwise separable convolutions. In the five open-source spectral data sets, PACE achieved a state-of-the-art performance. The more difficult the classification, the better the performance of PACE, compared with that of residual neural network (ResNet), vision transformer (ViT), and other commonly used deep learning algorithms. PACE helped improve the accuracy to 92.1% in Raman identification of pathogen-derived extracellular vesicles at different physiological states, which is much better than those of ResNet (85.1%) and ViT (86.0%). In general, the precise recognition and extraction of subtle differences offered by PACE are expected to facilitate vibrational spectroscopy to be a powerful tool toward revealing the relevant chemical reaction mechanisms in surface science or realizing the early diagnosis in life science.

Deep Learning-Enabled Raman Spectroscopic Identification of Pathogen-Derived Extracellular Vesicles and the Biogenesis Process.

Pathogenic bacterial infections, exacerbated by increasing antimicrobial resistance, pose a major threat to human health worldwide. Extracellular vesicles (EVs), secreted by bacteria and acting as their "long-distance weapons", play an important role in the occurrence and development of infectious diseases. However, no efficient methods to rapidly detect and identify EVs of different bacterial origins are available. Here, label-free Raman spectroscopy in combination with a new deep learning model of the attentional neural network (aNN) was developed to identify pathogen-derived EVs at Gram±, species, strain, and even down to physiological levels. By training the aNN model with a large Raman data set from six typical pathogen-derived EVs, we achieved the identification of EVs with high accuracies at all levels: exceeding 96% at the Gram and species levels, 93% at the antibiotic-resistant and sensitive strain levels, and still above 87% at the physiological level. aNN enabled Raman spectroscopy to interrogate the bacterial origin of EVs to a much higher level than previous methods. Moreover, spectral markers underpinning EV discrimination were uncovered from subtly different EV spectra via an interpretation algorithm of the integrated gradient. A further comparative analysis of the rich Raman biochemical signatures of EVs and parental pathogens clearly revealed the biogenesis process of EVs, including the selective encapsulation of biocomponents and distinct membrane compositions from the original bacteria. This developed platform provides an accurate and versatile means to identify pathogen-derived EVs, spectral markers, and the biogenesis process. It will promote rapid diagnosis and allow the timely treatment of bacterial infections.

[Reproducibility and repeatability of the determination of occlusal plane on digital dental models].

OBJECTIVE: To assess the repeatability(intraobserver comparison)and reproducibility(interobserver comparison)of two different methods for establishing the occlusal plane on digital dental models. METHODS: With Angle's classification as a stratification factor,48 cases were randomly extracted from 806 ones which had integrated clinical data and had their orthodontic treatment from July 2004 to August 2008 in Department of Orthodontics ,Peking University School and Hospital of Stomatology.Post-treatment plaster casts of 48 cases were scanned by Roland LPX-1200 3D laser scanner to generate geometry data as research subjects.In a locally developed software package,one observer repeated 5 times at intervals of at least one week to localize prescriptive landmarks on each digital model to establish a group of functional occlusal planes and a group of anatomic occlusal planes, while 6 observers established two other groups of functional and anatomic occlusal planes independently.Standard deviations of dihedral angles of each group on each model were calculated and compared between the related groups.The models with the five largest standard deviations of each group were studied to explore possible factors that might influence the identification of the landmarks on the digital models. RESULTS: Significant difference of intraobserver variability was not detected between the functional occlusal plane and the anatomic occlusal plane (P>0.1), while that of interobserver variability was detected (P<0.001). The pooled experimental standard deviation the 6 observers established for the functional occlusal plane was 0.2° smaller than that of the anatomic occlusal plane.The functional occlusal plane's variability of intraobserver and interobsever did not differ significantly (P>0.1), while the anatomic occlusal plane's variability of the intraobserver was significantly smaller than that of the interobserver (P<0.001), with a 0.12° difference value of the pooled experimental standard deviation. CONCLUSION: When one observer measures a number of digital models,both the functional and the anatomic occlusal planes are suitable as a conference plane with equal repeatability. When several observers measure a large number of digital models,the functional occlusal plane is more reproducible than the anatomic occlusal plane ,but the difference is small.

Effects of soil protists on the antibiotic resistome under long term fertilization.

Soil protists are key in regulating soil microbial communities. However, our understanding on the role of soil protists in shaping antibiotic resistome is limited. Here, we considered the diversity and composition of bacteria, fungi and protists in arable soils collected from a long-term field experiment with multiple fertilization treatments. We explored the effects of soil protists on antibiotic resistome using high-throughput qPCR. Our results showed that long term fertilization had stronger effect on the composition of protists than those of bacteria and fungi. The detected number and relative abundance of antibiotic resistance genes (ARGs) were elevated in soils amended with organic fertilizer. Co-occurrence network analysis revealed that changes in protists may contribute to the changes in ARGs composition, and the application of different fertilizers altered the communities of protistan consumers, suggesting that effects of protistan communities on ARGs might be altered by the top-down impact on bacterial composition. This study demonstrates soil protists as promising agents in monitoring and regulating ecological risk of antibiotic resistome associated with organic fertilizers.

[MiR-135b promotes proliferation of endometrial carcinoma cells by targeting FOXO1].

OBJECTIVE: To explore the expression of miR-135b in endometrial carcinoma and the mechanism by which miR-135b promotes the proliferation of endometrial cancer cells. METHODS: The expressions of miR-135b and FOXO1 were using RT-PCR detected in 22 fresh endometrial cancer tissues and paired adjacent tissues and also in endometrial cancer cell lines JEC, Ishikawa, HEC-1-B, and RL-952. The RL-952 and Ishikawa cell lines were transfected with miR-135b mimics or inhibitors, and the changes in their proliferative activity were detected with MTT assay; the expressions of FOXO1 mRNA and protein were detected by RT-PCR and Western blotting, respectively. RESULTS: The expression of miRNA135b was significantly up-regulated and FOXO1 expression was down-regulated in endometrial carcinoma tissues as compared with the adjacent tissues (P<0.05). The mRNA expression of miR-135b was negatively correlated with the expression of FOXO1 in endometrial carcinoma. In RL-952 and Ishikawa cell lines, transfection with miR-135b mimics obviously promoted the cell proliferation (P<0.05). Up-regulation of miR-135b significantly decreased the expressions of FOXO1 protein and mRNA (P<0.05), and down- regulation of miR-135b increased FOXO1 expressions (P<0.05). CONCLUSIONS: MiR-135b plays an important role in the occurrence and development of endometrial carcinoma partially by regulating its target gene FOXO1.

[Tip and torque changes in maxillary buccal segment after orthodontic treatment: a three-dimensional study].

OBJECTIVE: To investigate three-dimensional imaging registration and superimposition techniques in measuring the tip and torque change of upper canine, premolar and first molar after orthodontic treatment. METHODS: Twenty-eight subjects (14 extraction cases and 14 non-extraction cases) with full records were randomly selected from the Department of Orthodontics, Peking University School and Hospital of Stomatology. The pre-and post-treatment upper dental casts were digitized with three-dimensional spot laser scanner and superimposed with reverse engineering software. The facial axis of the clinical crown (FACC) was transferred from post-treatment teeth to the pre-treatment teeth using three-dimensional imaging registration. The occlusal plane was constructed on the post-treatment upper digital cast and the tip and torque values were measured. RESULTS: In the non-extraction group, the tip of the second premolar decreased by 1.5° (P < 0.05). The buccal crown torque of the first and second premolars increased by 5.1° and 4.2° (P < 0.05), respectively. In the extraction group, the lingual crown torque of upper canine increased by 3.8° (P < 0.05). CONCLUSIONS: Non-extraction orthodontic treatment tended to tip the upper second premolar distally and increased the buccal crown torque of the upper premolars while extraction treatment increased the lingual crown torque of the upper canine.

[Association of transforming growth factor-alpha gene polymorphism and environment factors with nonsyndromic cleft lip with or without cleft palate in Han nationality].

OBJECTIVE: To study the association of transforming growth factor-alpha (TGF-alpha gene polymorphism and environment factors with nonsyndromic cleft lip with or without cleft palate (NSCLP) in Han nationality. METHODS: Data related to infection, drug intake and folic acid supplement during pregnancy were gained through investigation of mothers. Polymerase chain reaction combined with restrict enzyme digestion was used to detect the target gene variation in 199 patients with NSCL/P and 203 healthy controls. Analysis was carried on the genotype and infection,drug intake and folic acid supplement. RESULTS: The C2 allele frequency in patients with NSCL/P was significantly higher than that in healthy controls. There was a significant increase of patients with NSCL/P in pregnant women exposed to infection, drug intake and folic acid deficiency. There was an interaction between C1C2 genetype and infection, drug intake and folic acid supplement. CONCLUSION: TGF-alpha gene polymorphism is associated with NSCL/P. Infection, drug intake and folic acid supplement during pregnancy were associated with the occurrence of NSCL/P. Individuals containing C2 allele were more sensitive to infection, drug intake and folic acid deficiency.