Predicting binding affinities of emerging variants of SARS-CoV-2 using spike protein sequencing data: observations, caveats and recommendations.

Predicting protein properties from amino acid sequences is an important problem in biology and pharmacology. Protein-protein interactions among SARS-CoV-2 spike protein, human receptors and antibodies are key determinants of the potency of this virus and its ability to evade the human immune response. As a rapidly evolving virus, SARS-CoV-2 has already developed into many variants with considerable variation in virulence among these variants. Utilizing the proteomic data of SARS-CoV-2 to predict its viral characteristics will, therefore, greatly aid in disease control and prevention. In this paper, we review and compare recent successful prediction methods based on long short-term memory (LSTM), transformer, convolutional neural network (CNN) and a similarity-based topological regression (TR) model and offer recommendations about appropriate predictive methodology depending on the similarity between training and test datasets. We compare the effectiveness of these models in predicting the binding affinity and expression of SARS-CoV-2 spike protein sequences. We also explore how effective these predictive methods are when trained on laboratory-created data and are tasked with predicting the binding affinity of the in-the-wild SARS-CoV-2 spike protein sequences obtained from the GISAID datasets. We observe that TR is a better method when the sample size is small and test protein sequences are sufficiently similar to the training sequence. However, when the training sample size is sufficiently large and prediction requires extrapolation, LSTM embedding and CNN-based predictive model show superior performance.

RSDNet: A New Multiscale Rail Surface Defect Detection Model.

The rapid and accurate identification of rail surface defects is critical to the maintenance and operational safety of the rail. For the problems of large-scale differences in rail surface defects and many small-scale defects, this paper proposes a rail surface defect detection algorithm, RSDNet (Rail Surface Defect Detection Net), with YOLOv8n as the baseline model. Firstly, the CDConv (Cascade Dilated Convolution) module is designed to realize multi-scale convolution by cascading the cavity convolution with different cavity rates. The CDConv is embedded into the backbone network to gather earlier defect local characteristics and contextual data. Secondly, the feature fusion method of Head is optimized based on BiFPN (Bi-directional Feature Pyramids Network) to fuse more layers of feature information and improve the utilization of original information. Finally, the EMA (Efficient Multi-Scale Attention) attention module is introduced to enhance the network's attention to defect information. The experiments are conducted on the RSDDs dataset, and the experimental results show that the RSDNet algorithm achieves a mAP of 95.4% for rail surface defect detection, which is 4.6% higher than the original YOLOv8n. This study provides an effective technical means for rail surface defect detection that has certain engineering applications.

The Key Role of Plant Hormone Signaling Transduction and Flavonoid Biosynthesis Pathways in the Response of Chinese Pine (Pinus tabuliformis) to Feeding Stimulation by Pine Caterpillar (Dendrolimus tabulaeformis).

In nature, plants have developed a series of resistance mechanisms to face various external stresses. As understanding of the molecular mechanisms underlying plant resistance continues to deepen, exploring endogenous resistance in plants has become a hot topic in this field. Despite the multitude of studies on plant-induced resistance, how plants respond to stress under natural conditions remains relatively unclear. To address this gap, we investigated Chinese pine (Pinus tabuliformis) using pine caterpillar (Dendrolimus tabulaeformis) under natural conditions. Healthy Chinese pine trees, approximately 10 years old, were selected for studying induced resistance in Huangtuliangzi Forestry, Pingquan City, Chengde City, Hebei Province, China. Pine needles were collected at 2 h and 8 h after feeding stimulation (FS) via 10 pine caterpillars and leaf clipping control (LCC), to simulate mechanical damage caused by insect chewing for the quantification of plant hormones and transcriptome and metabolome assays. The results show that the different modes of treatments significantly influence the contents of JA and SA in time following treatment. Three types of differentially accumulated metabolites (DAMs) were found to be involved in the initial response, namely phenolic acids, lipids, and flavonoids. Weighted gene co-expression network analysis indicated that 722 differentially expressed genes (DEGs) are positively related to feeding stimulation and the specific enriched pathways are plant hormone signal transduction and flavonoid biosynthesis, among others. Two TIFY transcription factors (PtTIFY54 and PtTIFY22) and a MYB transcription factor (PtMYB26) were found to be involved in the interaction between plant hormones, mainly in the context of JA signal transduction and flavonoid biosynthesis. The results of this study provide an insight into how JA activates, serving as a reference for understanding the molecular mechanisms of resistance formation in conifers responding to mandibulate insects.

Stochastic model for cell population dynamics quantifies homeostasis in colonic crypts and its disruption in early tumorigenesis.

The questions of how healthy colonic crypts maintain their size, and how homeostasis is disrupted by driver mutations, are central to understanding colorectal tumorigenesis. We propose a three-type stochastic branching process, which accounts for stem, transit-amplifying (TA) and fully differentiated (FD) cells, to model the dynamics of cell populations residing in colonic crypts. Our model is simple in its formulation, allowing us to estimate all but one of the model parameters from the literature. Fitting the single remaining parameter, we find that model results agree well with data from healthy human colonic crypts, capturing the considerable variance in population sizes observed experimentally. Importantly, our model predicts a steady-state population in healthy colonic crypts for relevant parameter values. We show that APC and KRAS mutations, the most significant early alterations leading to colorectal cancer, result in increased steady-state populations in mutated crypts, in agreement with experimental results. Finally, our model predicts a simple condition for unbounded growth of cells in a crypt, corresponding to colorectal malignancy. This is predicted to occur when the division rate of TA cells exceeds their differentiation rate, with implications for therapeutic cancer prevention strategies.

Waiting times in a branching process model of colorectal cancer initiation.

We study a multi-stage model for the development of colorectal cancer from initially healthy tissue. The model incorporates a complex sequence of driver gene alterations, some of which result in immediate growth advantage, while others have initially neutral effects. We derive analytic estimates for the sizes of premalignant subpopulations, and use these results to compute the waiting times to premalignant and malignant genotypes. This work contributes to the quantitative understanding of colorectal tumor evolution and the lifetime risk of colorectal cancer.

Climatic change in southern Kazakhstan since 1850 C.E. inferred from tree rings.

Although global warming is an indisputable fact, there is still uncertainty about how climate change will occur at regional levels. Kazakhstan is the largest landlocked country in the world. To best manage this country's limited water resources, socio-economic development and environmental protection, a solid understanding of regional climate change impacts is needed. In this study, tree-ring width and δ13C chronologies were established based on 99 tree-ring samples of Schrenk spruce (Picea schrenkiana Fisch. et Mey.) collected in Almaty, Kazakhstan. Climate response analysis between the tree-ring chronologies and climate data indicates that summer mean temperature is the strongest climate signal recorded by tree-ring δ13C. We reconstructed temperature change in southern Kazakhstan since 1850 C.E. using the tree-ring δ13Ccorr chronology. The results show that the temperatures in southern Kazakhstan have risen at a rate of about 0.27 °C per decade over the past 166 years. However, the rate has increased by as much as 0.44 °C per decade over the past 30 years. Analyses of temperature and precipitation data show that the climate has alternated between warm-dry and cold-humid periods over the past 166 years. The extreme droughts of 1879, 1917 and 1945 were caused by the combination of continuously high temperatures and reduced precipitation.

Tree growth response of Fokienia hodginsii to recent climate warming and drought in southwest China.

To date, few attempts have been made to assess the influence of climate change on forest ecosystems and on the relationship between tree growth and climate in humid areas of low latitudes. In this paper, we studied the response of tree growth and forest ecosystem to climate change by using Fokienia hodginsii tree-ring cores from the northern Yunnan-Guizhou Plateau, southwest of China. Tree growth correlates the highest (r = -0.64, p < 0.01) with mean temperature (July-September), but the coefficients were changing with time as revealed by a moving correlation analysis. Tree growth is significantly (p < 0.05) and positively correlated with January-April mean temperature from AD 1961-1987, while correlations with precipitation are insignificant. In contrast, from 1988 to 2014, tree growth correlated negatively with mean temperature of previous summer and positively with precipitation of previous August-September. This indicated that the limiting factors for tree growth have changed under different climate conditions. The meteorological data suggested that from 1961 to 1987 it was cold and wet in the study area and radial growth is limited by winter and spring temperatures. This restriction is weaker if the climate is appropriate in general. However, from 1988 to 2014, the combined effects of recent warming and decreasing precipitation have led to an increasing response of tree-ring width to drought. In addition, a large proportion of mature F. hodginsii mortality occurred from 2007 to 2013, which corresponds with a drastic reduction of radial growth (narrowest in recent 100 years). The recent drought, induced by decreasing precipitation and increasing temperature, may have passed the threshold which F. hodginsii could tolerate, causing tree growth reduction, tree growth-climate relationship change, as well as catastrophic tree mortality. All these changes may lead to further responses of the local ecosystem to climate change which should be highly regarded.

What Happens to LiMnPO4 upon Chemical Delithiation?

Olivine MnPO4 is the delithiated phase of the lithium-ion-battery cathode (positive electrode) material LiMnPO4, which is formed at the end of charge. This phase is metastable under ambient conditions and can only be produced by delithiation of LiMnPO4. We have revealed the manganese dissolution phenomenon during chemical delithiation of LiMnPO4, which causes amorphization of olivine MnPO4. The properties of crystalline MnPO4 obtained from carbon-coated LiMnPO4 and of the amorphous product resulting from delithiation of pure LiMnPO4 were studied and compared. The phosphorus-rich amorphous phases in the latter are considered to be MnHP2O7 and MnH2P2O7 from NMR, X-ray absorption spectroscopy, and X-ray photoelectron spectroscopy analysis. The thermal stability of MnPO4 is significantly higher under high vacuum than at ambient condition, which is shown to be related to surface water removal.

Topological regression as an interpretable and efficient tool for quantitative structure-activity relationship modeling.

Quantitative structure-activity relationship (QSAR) modeling is a powerful tool for drug discovery, yet the lack of interpretability of commonly used QSAR models hinders their application in molecular design. We propose a similarity-based regression framework, topological regression (TR), that offers a statistically grounded, computationally fast, and interpretable technique to predict drug responses. We compare the predictive performance of TR on 530 ChEMBL human target activity datasets against the predictive performance of deep-learning-based QSAR models. Our results suggest that our sparse TR model can achieve equal, if not better, performance than the deep learning-based QSAR models and provide better intuitive interpretation by extracting an approximate isometry between the chemical space of the drugs and their activity space.

Disruption of the Hippo pathway promotes the proliferation of childhood acute lymphoblastic leukemia cells, inhibits apoptosis and chemosensitivity.

BACKGROUND: Despite advancements in chemotherapy and stem cell transplantation, the recurrence and chemoresistance of childhood acute lymphoblastic leukemia (cALL) remain a significant challenge, thus indicating the need for novel therapeutic targets. RESEARCH DESIGN AND METHODS: The protein levels of YAP1, p-YAP1, TAZ, and Cyr61 of cALL patients and healthy volunteers were measured by western blot analysis. Then the leukemic cell line SUP-B15 was transfected with sh-YAP1 and pcDNA3.1-YAP1 to knockdown or overexpress YAP1. The viability, chemosensitivity, apoptosis, migration, and invasion of SUP-B15 cells were determined by MTT, flow cytometry, and Transwell assay. RESULTS: The cALL patients had higher YAP1, TAZ, and Cyr61 protein expression and lower p-YAP1 protein expression in bone marrow tissues compared with healthy volunteers (p < 0.01). In SUP-B15 cells, YAP1 knockdown upregulated p-YAP1 protein expression (p < 0.01) and downregulated TAZ and Cyr61 protein expression (p < 0.01). In addition, knocking down YAP1 significantly inhibited cell viability, migration, and invasion, and induced apoptosis (p < 0.01). YAP1 knockdown also reduced the IC50 value following treatment with vincristine, daunorubicin, cyclophosphamide, and dexamethasone (p < 0.05). CONCLUSIONS: Disruption of the Hippo pathway attenuates the development of cALL by promoting cell proliferation while suppressing apoptosis and drug sensitivity.

The quantification of circular RNA 0007841 during induction therapy helps estimate the response and survival benefits to bortezomib-based regimen in multiple myeloma.

OBJECTIVE: Circular RNA_0007841 (Circ_0007841) facilitates multiple myeloma (MM) progression and resistance of the bortezomib by experimental studies, while its clinical implication in MM patients is still unclear. This study intended to evaluate the longitudinal change and prognostic role of circ_0007841 expression in MM patients receiving bortezomib-based induction therapy. METHODS: In this prospective study, bone marrow plasma cell (BMPC) samples were gained from 97 MM patients at diagnosis and after bortezomib-based induction therapy, and from 30 healthy controls (HCs) proposing BM donation. Then, circ_0007841 expression in BMPC samples was measured by reverse transcription-quantitative polymerase chain reaction. Additionally, MM patients were followed up for a median of 29.4 months. RESULTS: Circ_0007841 expression was increased in MM patients compared to HCs (P < 0.001), but it was decreased after bortezomib-based induction therapy in MM patients (P < 0.001). Moreover, circ_0007841 expression at diagnosis was associated with the presence of t (4; 14) (P = 0.034), while its expression after bortezomib-based induction therapy was linked with higher revised international staging system stage (P = 0.025) in MM patients. Interestingly, circ_0007841 expression after bortezomib-based induction therapy was lower in MM patients who achieved complete remissions (P = 0.001) and overall responses (P = 0.002) compared to those who did not. Prognostically, circ_0007841 expression after bortezomib-based induction therapy (over the median vs. below the median) independently predicted shorter progression-free survival (hazard ratio (HR): 2.497, P = 0.002) and overall survival (HR: 3.107, P = 0.008) in MM patients. CONCLUSION: Circ_0007841 quantification during induction therapy may reflect the response and survival benefits to bortezomib-based regimen in MM patients.

Correlation analysis of bone marrow microvessel density and miRNA expression on drug resistance in patients with chronic myelogenous leukemia after tyrosine kinase inhibitor treatment.

OBJECTIVE: This study analyzed the relationship between bone marrow microvessel density (MVD) and the expression of four miRNAs with chronic myelogenous leukemia (CML) resistance after tyrosine kinase inhibitor (TKI) treatment. METHODS: 234 CML patients were divided into resistance and non-resistance groups in terms of the results of the 5-year follow-up. Patients were divided into the Optimum response group and the Warning/Failure group based on TKI response. MVD was determined by immunohistochemistry, and the expression levels of four miRNAs (miR-106a, miR-155, miR-146a, and miR-340) in bone marrow biopsy specimens were examined by qPCR. We evaluated the association of MVD with four miRNAs and them predictive value for CML resistance after TKI treatment. RESULTS: The MVD and the levels of miR-106a, miR-155, and miR-146a were significantly higher while the miR-340 level was lower in the resistance group than the non-resistance group. Besides, MVD had a significant correlation with the levels of miR-340 and miR-155. According to the results of survival analysis, MVD as well as miR-340 and miR-155 levels were observably correlated with 5-year survival of patients without TKI resistance. The results of the ROC curve indicated that the MVD, miR-106a, miR-340, and miR-155 had good predictive accuracy for CML resistance after TKI treatment. As for the results of multivariate analysis, disease stage, risk level (high risk), high MVD, low miR-340 expression, and high miR-155 expression were all independent risk factors for CML resistance. CONCLUSION: MVD and the expression of miR-340 and miR-155 are closely associated with CML resistance after TKI treatment.

[Growth decline characteristics of Picea schrenkiana at different altitudes in Yili River Basin, western Tian-shan Mountains, Xinjiang, China]. / 新疆天山西部伊犁河流域不同海拔雪岭云杉生长衰退特征.

Picea schrenkiana is the dominant tree species in Ili River Basin located in the western Tianshan Mountains of Xinjiang. We investigated the growth decline characteristics of P. schrenkiana at different altitudes (1800, 2300 and 2800 m) based on tree-ring index (TRI) and percentage growth change (GC), aiming to understand the growth response of P. schrenkiana to drought events at different altitudes and the impacts of altitude on tree growth decline in this region. The results showed that P. schrenkiana experienced multiple decline events at low-altitude (1800 m). TRI and GC identified inconsistent occurrence time of the decline events. The variations of TRI indicated that P. schrenkiana at low-altitude experienced two large-scale declines during 1927-1933 and 2017-2014, respectively. The variations of GC identified four decline events, including 1891-1893, 1924-1926, 1973-1975, and 2004-2009. The radial growth of P. schrenkiana across altitudes from low to high was significantly affected by the Palmer drought severity index (PDSI) of the previous growing season. The impact of current PDSI on P. schrenkiana during the growing season initially enhanced but later decreased with increasing altitude. In the extreme drought year 1917, the magnitude of growth decline increased with altitude. At low-altitude (1800 m), the TRI was 0.65, which was 35% lower than the normal level. At mid-altitude (2300 m) and high-altitude (2800 m), it was 0.56 and 0.54, respectively, being 40% lower than the average level. The drought event in 1917 had a 2-year legacy effect on the growth of P. schrenkiana at all the altitudes, with the TRI in 1920 recovered to exceeding 0.9, being close to the normal level. The impact of altitude on drought-induced forest decline was significant. Tree growth in low-altitude areas was more vulnerable to drought events due to the relatively poorer water and temperature conditions at low-altitude, which could lead to multiple large-scale decline events. In mid- and high-altitude areas, where hydrothermal conditions were more favorable, trees could experience even more severe decline during extreme droughts.

Protective Effect of the SIRT1-Mediated NF-κB Signaling Pathway against Necrotizing Enterocolitis in Neonatal Mice.

OBJECTIVE: To discover the mechanism of the sirtuin 1 (SIRT1)-mediated nuclear factor-κB (NF-κB) pathway in the protection against necrotizing enterocolitis (NEC) in neonatal mice. MATERIALS AND METHODS: Neonatal mice were treated with EX527 (an inhibitor of SIRT1) and/or pyrrolidine dithiocarbamate (PDTC, an inhibitor of NF-κB). The survival rate of the mice was recorded. Hematoxylin and eosin (HE) staining was performed to observe the pathological changes in the intestines. Furthermore, western blotting, enzyme-linked immunosorbent assay, and real-time quantitative polymerase chain reaction were conducted to measure the protein and gene expression, while corresponding kits were used to detect the levels of oxidative stress indicators. RESULTS: PDTC increased the survival rate of NEC mice. When compared with the NEC+ EX527 + PDTC group, the histological NEC score was higher in the NEC + EX527 group but lower in the NEC + PDTC group. SIRT1 expression in the intestines of NEC mice was downregulated, with an increase in p65 nuclear translocation. Additionally, malondialdehyde increased and glutathione peroxidase decreased in the intestines of NEC mice, with the upregulation of interleukin (IL)-6, IL-1ß, and tumor necrosis factor-α, as well as the downregulation of ZO-1, occludin, and claudin-4 in the intestines. However, the above changes could be improved by PDTC, which could be further reversed by EX527. CONCLUSION: SIRT1 can mitigate inflammation and the oxidative stress response and improve intestinal permeability by mediating the NF-κB pathway, playing an important role in the alleviation of NEC.

Tree-ring-based reconstruction of precipitation in the Changling Mountains, China, since A.D.1691.

Seven different tree-ring parameters were obtained from Chinese pine (Pinus tabulaeformis) in the Changling Mountains, China. The chronologies were analyzed individually and then compared with each other. The climate response analysis shows that total precipitation (September-July) is the main factor limiting the radial growth of Chinese pine in the Changling Mountains. Thus, the residual earlywood width chronology was used to estimate precipitation (September-July) for the period AD 1691-2006, and explained 46.9% of the precipitation variance. Drought events in our reconstruction are compared to historical archives for Gansu and north-central China. The results reveal the climatic extremes over much of Gansu. Some events have had profound impacts on the Gansu people over the past several centuries. Spatial analysis shows that the precipitation reconstruction has strong common signals for North-central China. The reconstructed series is correlated significantly with Helan Mountains Palmer drought severity index (PDSI), as well as with previous results from Jiuquan, Shandan, Huashan, Luya Mountains, and even the state of Mongolia. Our results suggest that some dry periods are coincident with solar minima over the past several hundred years. Multitaper spectral analysis reveals the existence of significant 24.4-year, 12.2-year, and 2.4- to 3.4-year periods of variability.

[Radial growth of dominant coniferous species and their responses to climate changes in the Altay Mountains, China]. / é˜¿å°”æ³°å±±ä¸»è¦é’ˆå¶æ ‘ç§æ ‘æœ¨å¾„å‘ç”Ÿé•¿åŠå ¶å¯¹æ°”å€™å˜åŒ–çš„å“åº”.

Based on the standard method of dendrochronology, we examined the tree-ring width index of two dominant tree species in the Altay Mountains, China, including Picea obovata and Larix sibirica. We analyzed the basal area increments (BAI) of those two species and the relationships between their radial growth and the climatic factors, which were compared in similar habitats. The results showed that the BAI of P. obovata was greater than L. sibirica, but the radial growth rate of L. sibirica was greater. In recent 60 years, the radial growth of P. obovata negatively correlated with high temperature in the fast growing stage of previous year, while the high temperature in June of current year promoted the radial growth of L. sibirica. There was a significantly negative correlation between radial growth of L. sibirica with temperature in January of current year. The sensitivity of tree growth to climate showed an obvious increase after an abrupt climate change under the background of recent warming and wetting trend in mid-1980s. Results of the moving correlation analysis showed that the response of the radial growth of P. obovata and L. sibirica to temperature and precipitation were enhanced under the background of climate change in the study area.

Discovery of a Potent Thieno[2,3-d]pyrimidine-2,4-dione-Based Protoporphyrinogen IX Oxidase Inhibitor through an In Silico Structure-Guided Optimization Approach.

A key objective for herbicide research is to develop new compounds with improved bioactivity. Protoporphyrinogen IX oxidase (PPO) is an essential target for herbicide discovery. Here, we report using an in silico structure-guided optimization approach of our previous lead compound 1 and designed and synthesized a new series of compounds 2-6. Systematic bioassays led to the discovery of a highly potent compound 6g, 1-methyl-3-(2,2,7-trifluoro-3-oxo-4-(prop-2-yn-1-yl)-3,4-dihydro-2H-benzo[b][1,4]oxazin-6-yl)thieno[2,3-d]pyrimidine-2,4(1H,3H)-dione, which exhibited an excellent and wide spectrum of weed control at the rates of 30-75 g ai/ha by the postemergence application and is relatively safe on maize at 75 g ai/ha. Additionally, the Ki value of 6g to Nicotiana tabacum PPO (NtPPO) was found to be 2.5 nM, showing 3-, 12-, and 18-fold higher potency relative to compound 1 (Ki = 7.4 nM), trifludimoxazin (Ki = 31 nM), and flumioxazin (Ki = 46 nM), respectively. Furthermore, molecular simulations further suggested that the thieno[2,3-d]pyrimidine-2,4-dione moiety of 6g could form a more favorable π-π stacking interaction with the Phe392 of NtPPO than the heterocyclic moiety of compound 1. This study provides an effective strategy to obtain enzyme inhibitors with improved performance through molecular simulation and structure-guided optimization.

Discovery of N-Phenylaminomethylthioacetylpyrimidine-2,4-diones as Protoporphyrinogen IX Oxidase Inhibitors through a Reaction Intermediate Derivation Approach.

Protoporphyrinogen oxidase (PPO, EC 1.3.3.4) is an effective target for green herbicide discovery. In this work, we reported the unexpected discovery of a novel series of N-phenylaminomethylthioacetylpyrimidine-2,4-diones (2-6) as promising PPO inhibitors based on investigating the reaction intermediates of our initially designed N-phenyluracil thiazolidinone (1). An efficient one-pot procedure that gave 41 target compounds in good to high yields was developed. Systematic Nicotiana tabacum PPO (NtPPO) inhibitory and herbicidal activity evaluations led to identifying some compounds with improved NtPPO inhibition potency than saflufenacil and good post-emergence herbicidal activity at 37.5-150 g of ai/ha. Among these analogues, ethyl 2-((((2-chloro-4-fluoro-5-(3-methyl-2,6-dioxo-4-(trifluoromethyl)-3,6-dihydropyrimidin-1(2H)-yl)phenyl)amino)methyl)thio)acetate (2c) (Ki = 11 nM), exhibited excellent weed control at 37.5-150 g of ai/ha and was safe for rice at 150 g of ai/ha, indicating that compound 2c has the potential to be developed as a new herbicide for weed management in paddy fields. Additionally, our molecular simulation and metabolism studies showed that the side chains of compound 2c could form a hydrogen-bond-mediated seven-membered ring system; substituting a methyl group at R1 could reinforce the hydrogen bond of the ring system and reduce the metabolic rate of target compounds in planta.

Characteristics of tree-ring density at different stem heights and their climatic responses.

Taking windfall woods of Picea schrenkiana in the southern mountainous area of the Ili Prefecture as the research object, tree-ring density chronologies were developed from the discs for maximum density (MXD), minimum density (MID), mean earlywood density (EWD), and mean latewood density (LWD) at five different stem heights (1.3, 5, 10, 15 and 20 m) to examine the climatic responses of tree-ring density by correlation analysis with local meteorological data. The results showed that there was a good coherence among the four types of tree-ring density chronologies for the same stem height, which was relatively significant for the data from 10, 15 and 20 m. The LWD had good coherence among different stem heights, while the climatic responses of tree-ring density at different stem heights varied. The MXD and LWD at 15 m were sensitive to mean tempera-ture from July to September in the previous year and from May to September in the current year. It might underestimate the response of P. schrenkiana to temperature if we sample tree-ring at 1.3 m.

Representation of features as images with neighborhood dependencies for compatibility with convolutional neural networks.

Deep learning with Convolutional Neural Networks has shown great promise in image-based classification and enhancement but is often unsuitable for predictive modeling using features without spatial correlations. We present a feature representation approach termed REFINED (REpresentation of Features as Images with NEighborhood Dependencies) to arrange high-dimensional vectors in a compact image form conducible for CNN-based deep learning. We consider the similarities between features to generate a concise feature map in the form of a two-dimensional image by minimizing the pairwise distance values following a Bayesian Metric Multidimensional Scaling Approach. We hypothesize that this approach enables embedded feature extraction and, integrated with CNN-based deep learning, can boost the predictive accuracy. We illustrate the superior predictive capabilities of the proposed framework as compared to state-of-the-art methodologies in drug sensitivity prediction scenarios using synthetic datasets, drug chemical descriptors as predictors from NCI60, and both transcriptomic information and drug descriptors as predictors from GDSC.