Breast cancer is a neoplastic disease which seriously threatens women’s health. It is regard as the most common cause of cancer death in women. Accurate detection and effective treatment are of vital significance to lower the death rate of breast cancer. In recent years, machine learning technique has been considered to be an effective method for accurate diagnosis of various diseases, among which Random Forest (RF) has been widely applied. However, decision trees with poor classification performance and high similarity may be generated during the training process, which affects the overall classification performance of the model. In this paper, a Hierarchical Clustering Random Forest (HCRF) model is developed. By measuring the similarity among all the decision trees, the hierarchical clustering technique is used to carry out clustering analysis on decision trees. The representative trees are selected from divided clusters to construct the hierarchical clustering random forest with low similarity and high accuracy. In addition, we use Variable Importance Measure (VIM) method to optimize the selected feature number for the breast cancer prediction. Wisconsin Diagnosis Breast Cancer (WDBC) database and Wisconsin Breast Cancer (WBC) database from the UCI (University of California Irvine) Machine Learning repository are employed in this study. The performance of the proposed method is evaluated by utilizing accuracy, precision, sensitivity, specificity and AUC (Area Under ROC Curve). Experimental results indicate that the classification based on HCRF algorithm with VIM as a feature selection method reaches the best accuracy of 97.05% and 97.76% compared to Decision Tree, Adaboost and Random Forest on both the WDBC and WBC datasets. The method proposed in this study is an effective tool for diagnosing breast cancer.

Small-sized DNA fragments were converted into relatively large DNA tetrahedrons based on the CRISPR-Cas12a technology. This conversion approach enlarged the signal of small targets, make it easier and more reproducible for glassy nanopores to detect.

Solid-state nanopores have been proven as a powerful platform for label-free single-molecule analysis. However, due to its relatively low resolution and selectivity, developing biosensors with good translocation signals faces two significant challenges: (1) small-sized chemical or biological targets show difficulty in producing recognizable translocation signals because of their weak interaction with the nanopore and (2) protein interferents that widely exist in biological samples or buffers would considerably deteriorate the noise level of the nanopore, submerging the translocation signal. Herein, we demonstrate an effective way to overcome both the challenges. DNA cubes were used as signal transducers that can achieve an ultra-high (>50 : 1) signal-to-noise ratio (SNR) translocation signal, which is maintained even in protein interferent-rich buffers. A sensing strategy was constructed via hepatitis B virus (HBV) target-triggered cleavage of the component elements of the DNA cube with the assistance of the CRISPR–Cas12a technology, which caused a great drop in the translocation rate. The elements to cleave were optimized, and the sensor performance was tested in different protein stabilizer-rich buffers and human serum. Coupling with the polymerase chain reaction (PCR) pre-amplification technology, HBV-positive or -negative classification was achieved with the detection limit reaching 5 aM. It is worth noting that in our method, all reaction buffers were directly used without further optimization, which is of great help for the practical application of solid-state nanopores.

Dibutyl phthalate (DBP) is a widely used plasticizer which has been found to be a reproductive and developmental toxicant and ubiquitously existing in the air. A highly sensitive method for DBP monitoring in the environment is urgently needed. A DBP sensor based on a homemade wireless-electrodeless quartz crystal microbalance with dissipation (QCM-D) coated with nano-structured nickel hydroxide is presented. With the noncontact configuration, the sensing system could work at a higher resonance frequency (the 3rd overtone) and the response of the system was even more stable compared with a conventional quartz crystal microbalance (QCM). The sensor achieved a sensitivity of 7.3 Hz/ppb to DBP in a concentration range of 0.4–40 ppb and an ultra-low detection limit of 0.4 ppb of DBP has also been achieved.

A novel design of dsDNA block contained a programmable sensing site inside is proposed, which can be programmed with different sequences to respond to multiple target molecules (e.g. Pb²⁺ and cfDNA) by being cleaved into two smaller DNA blocks.