Improving biosensor accuracy and speed using dynamic signal change and theory-guided deep learning.

False results and time delay are longstanding challenges in biosensing. While classification models and deep learning may provide new opportunities for improving biosensor performance, such as measurement confidence and speed, it remains a challenge to ensure that predictions are explainable and consistent with domain knowledge. Here, we show that consistency of deep learning classification model predictions with domain knowledge in biosensing can be achieved by cost function supervision and enables rapid and accurate biosensing using the biosensor dynamic response. The impact and utility of the methodology were validated by rapid and accurate quantification of microRNA (let-7a) across the nanomolar (nM) to femtomolar (fM) concentration range using the dynamic response of cantilever biosensors. Data augmentation and cost function supervision based on the consistency of model predictions and experimental observations with the theory of surface-based biosensors improved the F1 score, precision, and recall of a recurrent neural network (RNN) classifier by an average of 13.8%. The theory-guided RNN (TGRNN) classifier enabled quantification of target analyte concentration and false results with an average prediction accuracy, precision, and recall of 98.5% using the initial transient or entire dynamic response, which is indicative of high prediction accuracy and low probability of false-negative and false-positive results. Classification scores were used to establish new relationships among biosensor performance characteristics (e.g., measurement confidence) and design parameters (e.g., inputs and hyperparameters of classification models and data acquisition parameters) that may be used for characterizing biosensor performance.

Reduction of Biosensor False Responses and Time Delay Using Dynamic Response and Theory-Guided Machine Learning.

Here, we provide a new methodology for reducing false results and time delay of biosensors, which are barriers to industrial, healthcare, military, and consumer applications. We show that integrating machine learning with domain knowledge in biosensing can complement and improve the biosensor accuracy and speed relative to the performance achieved by traditional regression analysis of a standard curve based on the biosensor steady-state response. The methodology was validated by rapid and accurate quantification of microRNA across the nanomolar to femtomolar range using the dynamic response of cantilever biosensors. Theory-guided feature engineering improved the performance and efficiency of several classification models relative to the performance achieved using traditional feature engineering methods (TSFRESH). In addition to the entire dynamic response, the technique enabled rapid and accurate quantification of the target analyte concentration and false-positive and false-negative results using the initial transient response, thereby reducing the required data acquisition time (i.e., time delay). We show that model explainability can be achieved by combining theory-guided feature engineering and feature importance analysis. The performance of multiple classifiers using both TSFRESH- and theory-based features from the biosensor's initial transient response was similar to that achieved using the entire dynamic response with data augmentation. We also show that the methodology can guide design of experiments for high-performance biosensing applications, specifically, the selection of data acquisition parameters (e.g., time) based on potential application-dependent performance thresholds. This work provides an example of the opportunities for improving biosensor performance, such as reducing biosensor false results and time delay, using explainable machine learning models supervised by domain knowledge in biosensing.

Collaborative deep learning improves disease-related circRNA prediction based on multi-source functional information.

Emerging studies have shown that circular RNAs (circRNAs) are involved in a variety of biological processes and play a key role in disease diagnosing, treating and inferring. Although many methods, including traditional machine learning and deep learning, have been developed to predict associations between circRNAs and diseases, the biological function of circRNAs has not been fully exploited. Some methods have explored disease-related circRNAs based on different views, but how to efficiently use the multi-view data about circRNA is still not well studied. Therefore, we propose a computational model to predict potential circRNA-disease associations based on collaborative learning with circRNA multi-view functional annotations. First, we extract circRNA multi-view functional annotations and build circRNA association networks, respectively, to enable effective network fusion. Then, a collaborative deep learning framework for multi-view information is designed to get circRNA multi-source information features, which can make full use of the internal relationship among circRNA multi-view information. We build a network consisting of circRNAs and diseases by their functional similarity and extract the consistency description information of circRNAs and diseases. Last, we predict potential associations between circRNAs and diseases based on graph auto encoder. Our computational model has better performance in predicting candidate disease-related circRNAs than the existing ones. Furthermore, it shows the high practicability of the method that we use several common diseases as case studies to find some unknown circRNAs related to them. The experiments show that CLCDA can efficiently predict disease-related circRNAs and are helpful for the diagnosis and treatment of human disease.

Dual-targetable fluorescent probe for mapping the fluctuation of peroxynitrite in drug-induced liver injury model.

Drug-induced liver injury (DILI) is one of the most common and serious adverse drug reactions which can cause acute liver failure or even death in severe cases. With the incidence rate increasing over the years, DILI has became a frequent clinical liver disease and a global public health problem. As a biomarker of DILI, the detection of peroxynitrite (ONOO-) has became a powerful tool for the early diagnosis of liver injury. Here, we synthesized five mitochondria-targetable probes, 1-5, for detecting endogenous ONOO-. Through dye-screening, probe 5 was stood out by its excellent performance. In the presence of ONOO-, the fluorescence signal of probe 5 reduced 40-fold in 19 s with a low detection limit (9.36 nM). At the same time, the transformation can be observed with the naked eye under sunlight or UV lamp without being affected by the other reactive species. Even better, with low toxicity and high biocompatibility, probe 5 could successfully detect endogenous ONOO- in the mitochondrion of cells. Finally, probe 5 could specifically target the liver, and can be employed for monitoring the therapeutic effect of hepatoprotective medicine after drug-induced hepatotoxicity in vivo. In brief, probe 5 has the practical application capability for diagnosing the severity of the liver injury and researching the therapeutic effect of antidote in complex bio-systems.

Dual-Fluorophore and Dual-Site Multifunctional Fluorescence Sensor for Visualizing the Metabolic Process of GHS to SO2 and the SO2 Toxicological Mechanism by Two-Photon Imaging.

As a momentous gas signal molecule, sulfur dioxide (SO2) participates in diverse physiological activities. Excess SO2 will cause an apparent decrease in the level of intracellular glutathione (GSH), thereby damaging the body's antioxidant defense system. In addition, endogenous SO2 can be generated from GSH by reacting with thiosulfate (S2O32-) and enzymatically reduced to cysteine (Cys), a synthetic precursor of GSH. In view of their close correlation, a two-photon (TP) mitochondria-targeted multifunctional fluorescence sensor Mito-Na-BP was rationally designed and synthesized for detecting SO2 and GSH simultaneously. Under single-wavelength excitation, the sensor responded to GSH-SO2 and SO2-GSH continuously with blue-shifted and green fluorescence-enhanced signal modes, respectively, not just to GSH (enhanced) and SO2 (quenched) at 638 nm with a completely converse response tendency. Given its favorable spectral performance (high sensitivity, superior selectivity, and fast response rate) at physiological pH, Mito-Na-BP has been successfully applied in monitoring the level fluctuation of GSH affected from high-dose SO2 and visualizing in real time the metabolic process of GSH to SO2 by TP imaging. It is expected that this research will provide a convenient and efficient tool for elucidating intricate relationships of GSH and SO2 and facilitate further exploration of their functions in biomedicine.

Dual-Site Fluorescent Sensor as a Multiple Logic System for Studying the Dichotomous Function of Sulfur Dioxide under Oxidative Stress Induced by Peroxynitrite.

Intracellular reactive oxygen species and reactive sulfur play a vital role in regulating redox homeostasis and maintaining cell functions. Sulfur dioxide (SO2) has emerged as an important gas signal molecule recently, which is not only a potential reducing agent but also a potential inductor of oxidative stress in organisms. Due to high reactivity, peroxynitrite (ONOO-) could act on many biomolecules, such as proteins, lipids, and nucleic acids, and cause irreversible damage, eventually leading to cell apoptosis or necrosis. In order to further illuminate the dichotomous role of SO2 under oxidative stress induced by ONOO-, we designed the first dual-site fluorescent sensor (NIR-GYf) for separate or continuous detection of SO2 and ONOO-. NIR-GYf was successfully used for cell imaging of endogenous SO2 and ONOO-. In addition, western blotting analysis was used to verify the oxidation and antioxidation of SO2 and its dichotomous biological influence. Finally, NIR-GYf was integrated with multiple Boolean logic operations to construct an advanced analysis device, thereby realizing the direct analysis of SO2 and ONOO- levels.

Fluorescence distinguishing of SO2 derivatives and Cys/GSH from multi-channel signal patterns and visual sensing based on smartphone in living cells and environment.

Sulfur dioxide (SO2), cysteine (Cys) and glutathione (GSH), which perform crucial actions in regulating the balance of human, are closely related reactive sulfur species (RSS). Moreover, SO2 is one of the most concerned air pollutants, which is easily soluble in water and forms its derivatives. Therefore, it is highly desirable to differentiate SO2 derivatives and Cys/GSH in living cells and environment. Herein, a new near-infrared (NIR) mitochondria-targeted fluorescent probe, NIR-CG, which could distinguish SO2 derivatives and Cys/GSH by using multiple sets of signal patterns under single excitation was reported. NIR-CG exhibited different fluorescence signal modes to SO32- and Cys/GSH with low limit of detection (17.1 nM for SO32-, 17.3 nM for Cys and 25.9 nM for GSH). The recognition mechanisms of NIR-CG to SO32- and Cys/GSH were verified by HRMS, 1H NMR and DFT calculation. NIR-CG had good ability of mitochondrial targeted and fluorescence imaging in cells. What's more, NIR-CG showed great recovery rates (101-104%) in the determination of SO32- in actual water samples. It was worth noting that NIR-CG-based paper strip successfully realized the visual quantitative detection of SO32- and Cys/GSH by use of smartphone, which offered a novel method to develop powerful sensing platform.

A multifunctional fluorescent probe for visualizing H2S in wastewater with portable smartphone via fluorescent paper strip and sensing GSH in vivo.

In this paper, a bifunctional tri-site fluorescent probe was designed for the first time not only for visualization and quantitative analysis of sensing H2S in wastewater by coupling paper strip and smartphone (Color recognizer, Xiyi Technology) but also for sensitively monitoring GSH in living cells, which relied on different emission channels and the pH of solutions. The recognition properties of GH towards H2S/GSH were satisfactorily demonstrated through fluorescence, UV-vis, 1H NMR and DFT calculations. More importantly, integrated with the paper strip, portable smartphone-sensing platform with a color recognizer app would accomplish cost-effective and rapid assays for colorimetric water quality testing, which displayed huge application potential in fields of environmental monitoring.

A simple highly selective probe for discriminative visualization of endogenous cysteine, homocysteine and glutathione in living cells via three separated fluorescence channels.

Biothiols, as basic biological reactive sulfur species, perform vital actions in many critical physiological processes. Simultaneous detection and direct visualization of intracellular biothiols has great significance to figure out their metabolic mechanisms and cellular functions in living organism. However, it is an enormous challenge to selectively sense biothiols, troubled by their similar chemical structures and properties. In this work, we reported a novel chlorinated coumarin based multi-signal fluorescent probe (CC) for discriminative detection of cysteine (Cys), homocysteine (Hcy) and glutathione (GSH) through three different emission channels. CC exhibited favorable sensing properties towards biothiols such as large fluorescence enhancement, low limits of detection (6, 3 and 200 nM for Cys, Hcy and GSH), fast response speed (15 min) and pretty good water solubility. Moreover, it was also utilized for discriminative visualization of endogenous and exogenous Cys, Hcy and GSH in A375 cells and Tcam-2 cells successfully.

An ultrasensitive and visible lighting-up probe for imaging thiophenols in water samples, in serum and visualizing thiophenols-induced oxidative stress process in live cells.

Thiophenols, a class of significant industrial materials, are extremely toxic in environmental as well live cells. However, the process of live cells responding to thiophenols is not well understood. Herein, an innovative "OFF-ON" probe FY for thiophenols selectively in 100% aqueous solution was reported. It featured rapid response (~150 s), prominent sensitivity (detection limit: 5 nM), and large Stokes shift (~104 nm), which assured specific detection of thiophenols in A375 cells, HeLa cells and environment. Especially, it proved that thiophenols in live cells can be eliminated by endogenous reactive oxygen species (ROS), indicating that thiophenols may result in cellular oxidative stress. As well, it was resoundingly put into recognizing of thiophenols quantitatively in actual water samples and in serum.

Three isostructural Eu3+/Tb3+ co-doped MOFs for wide-range ratiometric temperature sensing.

Temperature is an important physical parameter in the fields of human life, scientific research and industrial applications, which requires accurate measurements in physiology and pathology. The fast, accurate, non-invasive and non-touch detection of temperature is one of the most important research findings of lanthanide MOFs (LnMOFs). Herein, three dual-emitting LnMOFs hybrids Tb1-xEuxL (x = 0.05, 0.1 and 0.2) were fabricated as ratiometric MOF thermometers. In the binary co-doped system, the special energy transfer method (Tb3+ → Eu3+) makes thermometers more sensitive and accurate. The detection range and maximum relative sensitivity (Sm) were investigated and compared with the reportorial materials. Luminescence explorations show that Tb0.9Eu0.1L can be used for sensing temperature in the range of 303-423 K and with a Sm of 1.75% K-1 at 423 K. Furthermore, PXRD experiment indicates that Tb0.9Eu0.1L has outstanding stability under simulated physiological environment, which makes the application of Tb0.9Eu0.1L to the detection of temperature in biomedical systems a reality.

Light-driven visualization of endogenous cysteine, homocysteine, and glutathione using a near-infrared fluorescent probe.

Herein, we presented a hydrosoluble triple-site and triple-excitation alternative NIR fluorescent probe for visualization of endogenous biothiols in phosphate-buffered saline (pH 7.4, 10 mM). Upon irradiation using different excitation light, probe 1 exhibited different fluorescence responses upon the addition of Cys, Hcy, and GSH: λex = 419 nm, λem = 498 nm; λex = 518 nm, λem = 573, 616, 727, and 783 nm; λex = 555 nm, λem = 612 and 727 nm, respectively. Furthermore, 1 was favourably applied for bioimaging endogenous Cys, Hcy, and GSH in A375 cells through well-defined blue-green-red emission channels.

Mito-Specific Ratiometric Terbium(III)-Complex-Based Luminescent Probe for Accurate Detection of Endogenous Peroxynitrite by Time-Resolved Luminescence Assay.

The level of peroxynitrite (ONOO-) is closely related to the pathogenesis of a series of diseases. Hence, the accurate sensing of ONOO- is very important for a better understanding of its biological functions. Here, a time-resolved ratiometric nanoprobe (TD-DC) composed of an energy donor unit (Tb(DPA)3, where DPA signifies a dipicolinate dianion ligand) and an energy receptor unit (DC) was designed. The response of TD-DC toward ONOO- in vitro was assessed by luminescence intensity and lifetime. The results demonstrated that TD-DC could precisely (39 nM) detect ONOO- with high selectivity and efficiency (29 s). TD-DC was further successfully employed in sensing endogenous ONOO- in A-375 cells. Moreover, it could specifically localize in the mitochondrion, where endogenous ONOO- is mainly generated. The above results revealed that TD-DC is efficient and useful for real-time detection of ONOO- at subcellular levels.

Highly specific monitoring and imaging of endogenous and exogenous cysteine in living cells.

Cys is one of the important biothiols and its abnormal concentration may pose a threat to human health. Therefore, the monitoring of Cys in organisms is of great significance. GSH and Hcy, as the other two biothiols, have similar chemical structures and active sites to Cys. Consequently, developing fluorescent probes to independently detect Cys has become a challenging problem. Keeping this in mind, α-ß unsaturated ketone as a recognition group was integrated into the coumarin group skeleton to synthesize a fluorescent probe SC. After the nucleophilic addition reaction of Cys with SC, the conjugated system of SC was blocked and the fluorescent enhanced obviously. SC was able to detect Cys specifically under the same excitation with a low detection limit (11.1 nM). SC showed a rapid respond to Cys (120 s) and good fluorescent stability over a wide pH range. In addition, it achieved extracorporeal circulation in the presence of H2O2 or NEM. In the end, SC could be applied to detecting endogenous and exogenous Cys under biological condition due to its slight cytotoxicity and good biocompatibility. This provided a powerful tool for studying the physiological function of Cys exclusively.

A coumarin-based colorimetric fluorescent probe for rapid response and highly sensitive detection of hydrogen sulfide in living cells.

Hydrogen sulfide (H2S) plays a vital role in numerous biological processes in living organisms. To better understand its functions, a fluorescent probe to fast and sensitively detect H2S is imminently needed. Keep this in mind, we reasonably designed probe DHC for detecting H2S based on α, ß-unsaturated ethanoylcoumarin fluorophore. The limit of detection (LOD) is found to be as low as 5 × 10-8 M, which is superior to most reported fluorescent probes to detect H2S. Furthermore, the wide pH range of 4-11 makes it capable of application in biological systems. Most importantly, MTT assays and cell imaging experiments indicate that probe DHC has hypotoxicity and outstanding membrane permeability, which makes DHC successful imaging of H2S in Baby Hamster Syrian Kidney (BHK) cells.

Dual-Site and Dual-Excitation Fluorescent Probe That Can Be Tuned for Discriminative Detection of Cysteine, Homocystein, and Thiophenols.

Thiols play a vital role in both the physiological process and organic synthesis field, including aliphatic thiols (e.g., Cys, Hcy, and GSH) and thiophenols. As a result of the similarities of thiols in terms of molecular structure and chemical properties, it is difficult for conventional fluorescent probes to distinguish them, which hinders the progress of biological and pathological research. Keeping this in mind, a dual-site and dual-excitation fluorescent probe (YY) was designed to distinguish among Cys, Hcy, and thiophenols by three different reaction paths. When excited at 470 nm, YY only exhibits a fluorescence OFF-ON response toward thiophenols. However, when excited at 453 nm, YY not only displays a fluorescence OFF-ON response toward Hcy and thiophenols (λem = 499 and 561 nm) but also presents a two-stage fluorescence response toward Cys, which possesses a fluorescence OFF-ON response in the first stage (λem = 501 nm) and then a fluorescence ON-OFF response in the second stage (λem = 556 nm). This specific fluorescence response indicates that YY has ability to overcome the above-mentioned challenge to achieve discriminative detection of Cys, Hcy, and thiophenols qualitatively, which promotes the study of thiols in the fields of physiology and pathology. Furthermore, cell-imaging studies show that YY can be applied to the imaging of exogenous Cys, Hcy, and thiophenols through two different emission channels.

A family of mixed-lanthanide metal-organic framework thermometers in a wide temperature range.

By choosing 2-pyridin-4-yl-4,5-imidazoledicarboxylic acid (H3PIDC) as the first ligand and sodium oxalate (OX) as the ancillary ligand, a series of mixed-lanthanide metal-organic frameworks (M'LnMOFs) [Tb1-xEux(HPIDC)(ox)1/2H2O]·3H2O (x = 0 1, 0.01 2a, 0.03 2b, 0.05 2c, 0.08 2d, 0.1 2e, 0.3 2f, 0.5 2g, 1 3) have been successfully synthesized via hydrothermal reactions. 2a-2f can serve as ratiometric luminescent sensors for detecting temperature. In this co-doped system, 2d shows an excellent linear response relationship with temperature from 303 to 473 K and exhibits a maximum relative sensitivity (Sr) of 0.60% K-1 at 473 K. Furthermore, powder X-ray diffraction (PXRD) experiments indicate that 2d has excellent chemical stability under simulated physiological conditions and alkali-acid solutions with pH ranging from 4 to 11, which makes it suitable to be applied in the physiological environment.

A multifunctional and recyclable terbium(iii) coordination polymer: displaying highly selective and sensitive detection of Fe3+ and CrVI anions, and picric acid in aqueous media.

A new two-dimensional (2D) terbium(iii) coordination polymer (1) was successfully fabricated by a solvothermal reaction. Luminescence spectra measurements show that 1 can selectively and sensitively detect Fe3+, CrVI (CrO42-/Cr2O72-), and picric acid (PA) in aqueous solution. The limits of detection (LODs) of 1 toward Fe3+, CrVI, and PA are calculated to be 50, 100, and 50 nM, which are significantly lower than the U.S. Environmental Protection Agency's (USEPA) proposed concentrations in drinking water. More significantly, 1 can be easily recycled at least 5 times in the detection process of CrVI (CrO42-/Cr2O72-) and PA through simple processing. For all we know, 1 is the first multifunctional and recyclable Tb-CP, which displays highly selective and sensitive detection of Fe3+, CrVI (CrO42-/Cr2O72-), and PA in aqueous media.

Coherence times of precise depth controlled NV centers in diamond.

We investigated the depth dependence of coherence times of nitrogen-vacancy (NV) centers through precise depth control using oxidative etching at 580 °C in air. By successive nanoscale etching, NV centers could be brought close to the diamond surface step by step, which enabled us to track the evolution of the number of NV centers remaining in the chip and to study the depth dependence of coherence times of NV centers with diamond etching. Our results showed that the coherence times of NV centers declined rapidly with the depth reduction in the last about 22 nm before they finally disappeared, which revealed a critical depth for the influence of a rapid fluctuating surface spin bath. Moreover, by using the slow etching method combined with low-energy nitrogen implantation, NV centers with depths shallower than the initially implanted depths can be generated, which are preferred for detecting external spins with higher sensitivity.