Drug development encompasses multiple processes, wherein protein subcellular localization is essential. It promotes target identification, treatment development, and the design of drug delivery systems. In this research, a deep learning framework called LocPro is presented for predicting protein subcellular localization. Specifically, LocPro is unique in (a) combining protein representations from the pre-trained large language model (LLM) ESM2 and the expert-driven tool PROFEAT, (b) implementing a hybrid deep neural network architecture that integrates convolutional neural network (CNN), fully connected (FC) layer, and bidirectional long short-term memory (BiLSTM) blocks, and (c) developing a multi-label framework for predicting protein subcellular localization at multiple granularity levels. Additionally, a dataset was curated and divided using a homology-based strategy for training and validation. Comparative analyses show that LocPro outperforms existing methods in sequence-based multi-label protein subcellular localization prediction. The practical utility of this framework is further demonstrated through case studies on drug target subcellular localization. All in all, LocPro serves as a valuable complement to existing protein localization prediction tools. The web server is freely accessible at https://idrblab.org/LocPro/.

Obejective To explore whether it is possible to detect the 123I-prostate-specific membrane antigen(PSMA)radiation value of the puncture tissue during prostate biopsy to achieve real-time,rapid,and accurate identification of benign and malignant prostate tissues,so as to improve the current clinical biopsy strategy and achieve accurate diagnosis of prostate cancer during operation with fewer puncture needles.Method In this prospective,diagnostic trial,we included 29 patients with suspected prostate cancer.All patients underwent transperineal biopsy guided by ultrasound within 24 hours after injection of 123I-PSMA,a total of 435 punctures were performed.The radiation value of punctured tissue was measured in real-time with a gamma counter.Pearson test is used to correlate radiation value with histopathology.Result The median radiation value of prostate cancer tissue(1 906.50 cpm)was significantly higher than that of benign prostate tissue(415.00 cpm).The optimal cut-off value for distinguishing benign and malignant prostate tissues was 828.50 cpm.The median radiation value of clinically significant prostate cancer tissue(2 652.50 cpm)was significantly higher than that of clinically insignificant prostate cancer(1 386.00 cpm).The optimal cut-off value for distinguishing clinically significant and clinically insignificant prostate cancer tissues was 1 767.00 cpm.In additional,there was a significant positive correlation between the radiation value of puncture tissue and ISUP pathological grade(r=0.834).Conclusion It is preliminarily confirmed that detection of 123I-PSMA radiation value of prostate puncture tissue can realize real-time,rapid and accurate identification of benign and malignant prostate tissues during operation.

Drug development encompasses multiple processes,wherein protein subcellular localization is essential.It promotes target identification,treatment development,and the design of drug delivery systems.In this research,a deep learning framework called LocPro is presented for predicting protein subcellular localization.Specifically,LocPro is unique in(a)combining protein representations from the pre-trained large language model(LLM)ESM2 and the expert-driven tool PROFEAT,(b)implementing a hybrid deep neural network architecture that integrates convolutional neural network(CNN),fully connected(FC)layer,and bidirectional long short-term memory(BiLSTM)blocks,and(c)developing a multi-label framework for predicting protein subcellular localization at multiple granularity levels.Additionally,a dataset was curated and divided using a homology-based strategy for training and validation.Compar-ative analyses show that LocPro outperforms existing methods in sequence-based multi-label protein subcellular localization prediction.The practical utility of this framework is further demonstrated through case studies on drug target subcellular localization.All in all,LocPro serves as a valuable complement to existing protein localization prediction tools.The web server is freely accessible at https://idrblab.org/LocPro/.

Obejective To explore whether it is possible to detect the 123I-prostate-specific membrane antigen(PSMA)radiation value of the puncture tissue during prostate biopsy to achieve real-time,rapid,and accurate identification of benign and malignant prostate tissues,so as to improve the current clinical biopsy strategy and achieve accurate diagnosis of prostate cancer during operation with fewer puncture needles.Method In this prospective,diagnostic trial,we included 29 patients with suspected prostate cancer.All patients underwent transperineal biopsy guided by ultrasound within 24 hours after injection of 123I-PSMA,a total of 435 punctures were performed.The radiation value of punctured tissue was measured in real-time with a gamma counter.Pearson test is used to correlate radiation value with histopathology.Result The median radiation value of prostate cancer tissue(1 906.50 cpm)was significantly higher than that of benign prostate tissue(415.00 cpm).The optimal cut-off value for distinguishing benign and malignant prostate tissues was 828.50 cpm.The median radiation value of clinically significant prostate cancer tissue(2 652.50 cpm)was significantly higher than that of clinically insignificant prostate cancer(1 386.00 cpm).The optimal cut-off value for distinguishing clinically significant and clinically insignificant prostate cancer tissues was 1 767.00 cpm.In additional,there was a significant positive correlation between the radiation value of puncture tissue and ISUP pathological grade(r=0.834).Conclusion It is preliminarily confirmed that detection of 123I-PSMA radiation value of prostate puncture tissue can realize real-time,rapid and accurate identification of benign and malignant prostate tissues during operation.

[Objective] To analyze the correlation of the distribution and morphological characteristics of breast calcification with the risk of breast malignant tumors, so as to further reveal the important value of the calcification in diagnosis of breast malignant tumors. [Methods] A total of 108 patients who had received concurrent puncture biopsy and surgical treatment of breast calcification at The Second Affiliated Hospital of Xi’an Jiaotong University from January 1, 2019 to December 31, 2019 were recruited, and their imaging data and postoperative pathological results were analyzed. We divided the distribution of breast X calcification into two categories based on segmental/linear calcification and regional/diffuse calcification. We classified the morphology of breast X calcification into punctate, small branching, polymorphic, amorphous, and coarse ones. We then explored the distribution, morphology, and BI-RADS classification of tumor calcification under breast X-ray, and its correlation with the occurrence, age, menopausal status, hormone receptors, and molecular typing of breast malignant tumors. [Results] Among the 108 patients enrolled in the study, 21 cases were malignant and 87 cases were benign. The results showed that the distribution of calcification was segmental/linear (χ²=11.2, P=0.003) and the calcification morphology was of small branching calcification (χ²=9.3, P=0.046). The detection rate in malignant tumors was significantly higher than that in benign tumors. The distribution and morphology of calcification were not significantly correlated with ER, PR, HER2, Ki67, pathological type, or molecular typing (P>0.05). Logistic regression analysis showed that , calcification distribution was segmental/linear [OR=19.94(3.061-129.9), P=0.002], calcification morphology was small branching [OR=3.906(1.141-13.37), P=0.030], and B-RADS classification was 4 grade and above [OR=39.99(2.703-591.7), P=0.007], which were closely related to the occurrence of malignant tumors. [Conclusion] For patients with breast tumors in which calcification can be detected under mammography, their imaging characteristics are closely related to the occurrence of malignant tumors. The distribution of calcification is segmental/linear, and the morphology of calcification is small branching/amorphous, which is more likely to be a malignant tumor and closely related to the occurrence of breast cancer.

Objective To investigate the effects of emotions(subjective well-being,depressed effect,worry,and guilt)on cancer(colorectal cancer,hepatic cancer,thyroid cancer,lung cancer,and breast cancer).Methods Two-sample bi-directional Mendelian randomization(MR)method was adopted.All data were based on summary data from genome-wide association studies(GWAS).Inverse variance weighting(IVW)was used to generate the main results,and weighted median(WM)and MR-Egger methods were employed to calculate supplementary results.The outcome measure was odds ratio(OR),and sensitivity analysis was conducted.Results For depressed effect,a significant association with lung cancer(OR=1.005,95％CI:1.001-1.009,P=0.015)was found.For worry,a significant association with breast cancer(OR=1.199,95％CI:1.011-1.423,P=0.038)was observed.For guilt,a significant association with thyroid cancer(OR=2.083,95％CI:1.080-4.017,P=0.029)was identified.After removing all potentially pleiotropic SNPs detected by MR PRESSO,the association between worry and breast cancer showed no statistical difference(P=0.064),while the association between worry and colorectal cancer remained significant(OR=0.739,95％CI:0.571-0.956,P=0.021).No causal relationship was found between cancer and emotions.Conclusion There is a causal relationship between depression and increased lung cancer incidence,guilt and increased thyroid cancer incidence,as well as anxiety and decreased colorectal cancer incidence.

Objective To investigate predictive value of combined lung injury prediction score(LIPS)and acute physiology and chronic health evaluation Ⅱ(APACHE Ⅱ)scores in patients with severe traumatic brain injury(sTBI)complicated with acute lung injury(ALI).Methods Seventy-five sTBI patients admitted to Provincial Hospital of Anhui Medical University from January 2019 to December 2021 were retrospectively selected and divided into ALI group(n=24)and non-ALI group(n=51)according to whether they were complicated with ALI.Basic data,laboratory indicators,APACHE Ⅱ score,LIPS score and Glasgow coma scale(GCS)score were collected.Logistic regression was used to analyze the risk factors of patients with sTBI complicated with ALI,and predictive value of the evaluation index of the receiver operating characteristic(ROC)curve for sTBI complicated with ALI was drawn.Results APACHE Ⅱ score and LIPS score in ALI group were significantly higher than those in non-ALI group,GCS score and red cell volume distribution width were significantly lower than those in non-ALI group(P<0.05).Logistic regression analysis showed that APACHE Ⅱ score,LIPS score and GCS score were independent risk factors for sTBI complicated with ALI(P<0.05).ROC curve analysis showed that area under the curve(AUC)of LIPS score and APACHE Ⅱ score in the diagnosis of sTBI complicated with ALI were 0.869 and 0.754,respectively.The AUC was 0.916(95％CI:0.855-0.976),and the sensitivity and specificity were 83.4％and 84.3％,respectively.Conclusion LIPS score combined with APACHE Ⅱ score can effectively predict the risk of sTBI complicated with ALI.