OpenFold: retraining AlphaFold2 yields new insights into its learning mechanisms and capacity for generalization.

AlphaFold2 revolutionized structural biology with the ability to predict protein structures with exceptionally high accuracy. Its implementation, however, lacks the code and data required to train new models. These are necessary to (1) tackle new tasks, like protein-ligand complex structure prediction, (2) investigate the process by which the model learns and (3) assess the model's capacity to generalize to unseen regions of fold space. Here we report OpenFold, a fast, memory efficient and trainable implementation of AlphaFold2. We train OpenFold from scratch, matching the accuracy of AlphaFold2. Having established parity, we find that OpenFold is remarkably robust at generalizing even when the size and diversity of its training set is deliberately limited, including near-complete elisions of classes of secondary structure elements. By analyzing intermediate structures produced during training, we also gain insights into the hierarchical manner in which OpenFold learns to fold. In sum, our studies demonstrate the power and utility of OpenFold, which we believe will prove to be a crucial resource for the protein modeling community.

Diagnostic significance of cerebrospinal fluid flow cytometry in Chinese children with B lineage acute lymphoblastic leukemia.

BACKGROUND: Central nervous system leukemia (CNSL) is one of the major causes of the poor prognosis of childhood leukemia. We aimed to compare the sensitivity of cytomorphology (CM) and flow cytometry (FCM) in diagnosing CNSL, emphasizing the importance of FCM in the diagnosis process. METHODS: One-hundred-sixty-five children with newly diagnosed B-cell Acute Lymphoblastic Leukemia (B-cell ALL) were included in this study. Cerebrospinal fluid (CSF) samples were taken for routine CSF analysis, CM analysis, and FCM examination. Computed tomography scans and/or magnetic resonance imaging were performed at diagnosis. Patients with CNS2, CNS3, and traumatic lumbar puncture (TLP) at diagnosis received two additional courses of triple intrathecal injections during induction treatment. We compared the sensitivity of FCM and CM in the diagnosis of children with CNSL. RESULTS: One hundred and twenty-eight (77.58%) CSF samples were negative by either CM or FCM (CM-/FCM-), four (2.42%) were positive by both CM and FCM (CM+/FCM+), and thirty-three (20%) displayed a single positive finding by FCM (CM-/FCM+) (p = 0.044). By adding two intrathecal injections in the induction treatment, ten children with TLP+ had no CNS relapse, like those with TLP-. However, compared to CNS1 and TLP, the event-free survival (EFS) did not significantly improve in patients with CNS2 and CNS3. Moreover, CNSL status was associated with worse 3-year EFS (p < 0.05). CONCLUSIONS: We have validated that FCM is more accurate in stratifying the status of the CNS compared to CM analysis. However, to improve the EFS rate of childhood leukemia, it is necessary to combine CM examination, FCM, and cranial imaging for the early diagnosis of CNSL.

Concomitant Distal Tibia-Fibula Fractures Treated with Intramedullary Nailing, With or Without Fibular Fixation: A Meta-Analysis.

There is no consensus on whether a fibular fracture should be fixed when a concurrent extra-articular distal tibia fracture is managed with intramedullary nails. We evaluated the use of fibular fixation in a meta-analysis of randomized trials and observational studies. Two researchers independently assessed the quality of eligible studies and extracted the data. We analyzed 4 trials with a pooled sample of 283 patients (mean age, 24 to 43 years; 141 men), 94 who had undergone fibular fixation and 189 who had not. Two randomized trials assessed on the Cochrane risk-assessment criteria were determined to have a moderate risk of bias, and 2 retrospective cohort studies evaluated with the Newcastle-Ottawa Scale were considered to be high quality. Tibia malalignment at follow-up times ranging from 12 to 72 weeks was reported in 20% (19/94) of patients in the fibular-fixation group and 67% (126/189) of patients in the nonfixation group, indicating that fibular fixation was significantly associated with a lower risk of malalignment (risk ratio, 0.34; 95% confidence interval [CI] 0.13 to 0.92; p = .03). The groups did not differ in the proportion of patients with malalignment immediately after surgery, delayed union, nonunion, or infection. When distal extra-articular tibia-fibular fractures are treated with intramedullary nails, fibular fixation may decrease the risk of late malalignment. Further randomized controlled trials with higher quality are required to verify the result.

The statistics of epidemic transitions.

Emerging and re-emerging pathogens exhibit very complex dynamics, are hard to model and difficult to predict. Their dynamics might appear intractable. However, new statistical approaches-rooted in dynamical systems and the theory of stochastic processes-have yielded insight into the dynamics of emerging and re-emerging pathogens. We argue that these approaches may lead to new methods for predicting epidemics. This perspective views pathogen emergence and re-emergence as a "critical transition," and uses the concept of noisy dynamic bifurcation to understand the relationship between the system observables and the distance to this transition. Because the system dynamics exhibit characteristic fluctuations in response to perturbations for a system in the vicinity of a critical point, we propose this information may be harnessed to develop early warning signals. Specifically, the motion of perturbations slows as the system approaches the transition.

Oxidation of Microcystic-LR via the solar/chlorine process: Radical mechanism, pathways and toxicity assessment.

Microcystin-LR (MC-LR) is the most widely distributed and harmful variant toxins released by cyanobacteria, which poses potential threaten to people and aquatic animals when entering natural water. In our research, solar/chlorine process was comprehensively investigated to degrade and detoxify MC-LR. Under the chlorine concentration of 1.0 mg L-1, MC-LR (1.0 µM) was decreased by 96.7%, 26%, and 9% by solar/chlorine process, chlorination, and solar irradiation respectively. Quenching experiments confirmed that reactive chlorine species (RCS) and hydroxyl radical (HO) were the predominant reactive species in solar/chlorine process at neutral condition, and ozone was generated because of the participation of triplet-state oxygen (O(3P)). The respective contributions of each reactive species were calculated with the order as: RCS, HO, ozone, and solar irradiation. The presence of HCO3- and natural organic matter in water inhibited the degradation efficiency of MC-LR. Moreover, the transformation products of MC-LR generated during the solar/chlorine process were identified and a possible pathway was proposed. The hepatotoxicity of MC-LR and its transformation products was compared using protein phosphatase 2A. Our experimental results revealed that the concentration and hepatotoxicity of MC-LR both significantly decreased, and most products were not hepatoxic. Overall, the solar/chlorine process is a promising alternative technology to degrade MC-LR during eutrophication.

An Improved Imaging Algorithm for High-Resolution Spotlight SAR with Continuous PRI Variation Based on Modified Sinc Interpolation.

This paper focuses on an improved imaging algorithm for spotlight synthetic aperture radar (SAR) with continuous Pulse Repetition Interval (PRI) variation in extremely high-resolution. Conventional SAR systems are limited in that a wide swath cannot be achieved with a high azimuth resolution in the meantime. This limitation can be overcome by Pulse Repetition Frequency (PRF) variation in a SAR system. However, there are problems such as the ambiguities of point targets or extended targets caused by nonuniform sampling. A reconstructive method, Nonuniform Discrete Fourier Transform (NUDFT) has been presented in the current literature, but it is rather computationally expensive. In this paper, a modified sinc interpolation based on NUDFT is proposed, which is used to reconstruct the uniformly sampled echo in time domain. Since the interpolation kernel length is relatively short, it is more computationally efficient. Then, the two-step processing approach combined with the modified sinc interpolation is further presented, which has much better accuracy than that combined with the conventional sinc interpolation. Both the simulated data and the extracted GF-3 data experiment demonstrate the validity and accuracy of the proposed approach.

The prevalence and morphological characteristics of the knee anterolateral ligament in a Chinese population.

The anterolateral ligament, a distinct structure connecting the lateral femoral epicondyle to the anterolateral proximal tibia, is gaining attention because of its possible function in ensuring internal rotational stability of the tibia. To study the prevalence and precise anatomical characteristics of the anterolateral ligament and its relationship to adjacent structures in a Chinese population, a total of 20 amputated knee specimens were collected. The anterolateral regions of the knees underwent detailed surgical dissection, followed by precise measurement of the anterolateral ligament and its adjacent structures. Histological analysis of the anterolateral ligament was performed using hematoxylin and eosin (H&E) staining. A thin soft tissue deep to the iliotibial band running obliquely across the lateral fibula ligament and connecting the lateral head of the gastrocnemius with the tibia, termed the 'gastrocnemius-tibial ligament' or superficial layer of the anterolateral ligament, was observed in 18 of the 20 specimens, corresponding to a prevalence of 90%. Furthermore, a well-defined anterolateral ligament deep to the gastrocnemius-tibial ligament and distinct from the lateral fibula ligament was found in all 20 knees (prevalence, 100%). The independent gastrocnemius-tibial ligament and anterolateral ligament had separate femoral originations at the lateral head of the gastrocnemius and the lateral femoral epicondyle, and the same osseous tibial insertion at the midpoint between Gerdy's tubercle and the most lateral aspect of the fibular head. H&E staining showed that both the anterolateral ligament and gastrocnemius-tibial ligament were ligaments consisting of collagenous bundles. In the Chinese Han population, the gastrocnemius-tibial ligament and anterolateral ligament may form a complex at the anterolateral aspect of the knee, which is likely involved in ensuring the internal rotational stability of the tibia.

Instantaneous brain dynamics mapped to a continuous state space.

Measures of whole-brain activity, from techniques such as functional Magnetic Resonance Imaging, provide a means to observe the brain's dynamical operations. However, interpretation of whole-brain dynamics has been stymied by the inherently high-dimensional structure of brain activity. The present research addresses this challenge through a series of scale transformations in the spectral, spatial, and relational domains. Instantaneous multispectral dynamics are first developed from input data via a wavelet filter bank. Voxel-level signals are then projected onto a representative set of spatially independent components. The correlation distance over the instantaneous wavelet-ICA state vectors is a graph that may be embedded onto a lower-dimensional space to assist the interpretation of state-space dynamics. Applying this procedure to a large sample of resting-state and task-active data (acquired through the Human Connectome Project), we segment the empirical state space into a continuum of stimulus-dependent brain states. Upon observing the local neighborhood of brain-states adopted subsequent to each stimulus, we may conclude that resting brain activity includes brain states that are, at times, similar to those adopted during tasks, but that are at other times distinct from task-active brain states. As task-active brain states often populate a local neighborhood, back-projection of segments of the dynamical state space onto the brain's surface reveals the patterns of brain activity that support many experimentally-defined states.

Regular biennial cycles in epidemics caused by parametric resonance.

The interaction between nonlinearity and seasonal forcing in childhood infectious diseases often leads to multiyear cycles with large amplitude. Regular biennial cycles in particular were observed in measles reports throughout the world. The objective of this paper is to understand the mechanism of such biennial cycles, especially the conditions under which the large amplitude biennial oscillation might appear. It is proposed that such biennial cycles are caused by parametric resonance, which might occur when varying the parameter at a frequency close to twice the natural frequency of the system. The analysis is carried out by solving an SIR model semi-analytically using method of multiple scales (MMS). This analysis shows how parametric resonance occurs due to the interaction between nonlinearity and periodic forcing. Using the MMS solution, the boundary between the resonance region and the non-resonance region in the parameter space is obtained. The effects of different parameters on the triggering of parametric resonance are studied, such as transmission rate, recovery rate, birth rate and amplitude of seasonality. The effects of stochasticity on the onset of parametric resonance are also studied.

Biomechanically based Fu's subcutaneous needling treatment for senile knee osteoarthritis: protocol for a randomized controlled trial.

INTRODUCTION: Fu's subcutaneous needling (FSN) is a new type of acupuncture that uses subcutaneous tissue to oscillate from side to side to improve muscle pathology status and can be effective in treating Knee osteoarthritis. Nonetheless, whether the clinical effect is similar to that of most commonly used drugs is unclear. Thus, this study aims to determine the pain-relieving effect and improvement in the joint function of the FSN therapy by comparing it with that of a positive control drug (celecoxib). Furthermore, this clinical trial also aims to evaluate the effect of FSN on gait and lower limb muscle flexibility, which can further explore the scientific mechanisms of the FSN therapy. METHODS AND ANALYSIS: This study is a randomized, parallel-controlled, single-center prospective clinical study that includes 60 participants, with an FSN group (n = 30) and a drug group (n = 30). The Fu's subcutaneous needling (FSN) group undergo the FSN therapy 3 times a week for 2 weeks, while the drug group receives 0.2 g/day oral celecoxib for 2 weeks, with a follow-up period of 4 weeks after the completion of treatment. The primary outcome is the difference in the visual analog scale score after 2 weeks of treatment compared with baseline. The Western Ontario and McMaster Universities (WOMAC) Osteoarthritis Index, joint active range of motion test, three-dimensional gait analysis, and shear wave elastic imaging technology analysis in lower limb muscles are also performed to demonstrate clinical efficacy. ETHICS AND DISSEMINATION: The trial is performed following the Declaration of Helsinki. The study protocol and consent form have been approved by the Ethics Committee of Guangdong Provincial Hospital of Chinese Medicine. All patients will give informed consent before participation and the trial is initiated after approval. The results of this trial will be disseminated through publication in peer-reviewed journals. TRIAL REGISTRATION NUMBER: NCT06328153.

Association between the C-reactive protein/albumin ratio and mortality in older Japanese patients with dysphagia.

Background: C-reactive protein-to-albumin ratio (CRP/ALB) has been proven to represent a biomarker for predicting prognosis in many groups of patients with severe diseases. However, few studies have investigated the association between CRP/ALB and mortality in Japan older people with dysphagia patients. Objective: This retrospective cohort study aimed to assess the prognostic value of C-reactive protein/albumin ratio (CAR) in older Japanese patients with dysphagia. Methods: We analyzed data from 253 patients diagnosed with dysphagia at a single center between January 2014 and January 2017. Cox regression analysis was used to compare the mortality rates across the CAR tertiles. Subgroup analyses were conducted, and Kaplan-Meier curves were used to determine the median survival times. Results: The study included 154 female and 99 male patients, with a median age of 83 years. After adjusting for all covariates, the multivariable Cox regression analysis revealed a significant association between increasing CAR (HR = 1.19, 95% CI: 1.03-1.37, P = 0.022) and the risk of mortality. Compared to the reference group T1 (< 0.149), the adjusted hazard ratios for T2 (0.149-0.815) and T3 (> 0.815) were 1.75 (95% CI: 1.07-2.87, P = 0.027) and 2.15 (95% CI: 1.34-3.46, P = 0.002), respectively. Kaplan-Meier curves indicated median survival times of 864, 371, and 223 days for T1, T2, and T3, respectively. Conclusion: The C-reactive protein/albumin ratio was positively related to mortality in Japan older people with dysphagia patients. There was no interaction for the subgroup analysis. The result was stable.

Core-shell structured lignin-stabilized silver nanoprisms for colorimetric detection of sulfur ions.

Widespread occurrence of sulfides in domestic and industrial wastewater contributes to environmental pollution and poses risks to human health. Therefore, the development of highly selective, sensitive, and rapid sulfur ion (S2-) detection probes in aquatic ecosystems is of paramount importance. In this study, lignin-stabilized silver nanoprisms (EHL@AgNPRs) were prepared using the seed growth and self-assembly methods. Based on this, a novel, high-performance, and environmentally friendly S2- colorimetric detection method was proposed. Lignin is believed to coat the surface of AgNPRs through cation-π and electrostatic interactions, acting as an excellent dispersant and stabilizer to prevent aggregation and shape deformation. This allows AgNPRs to maintain localized surface plasmon resonance (LSPR) characteristics and superior colorimetric sensing sensitivity towards S2- even after 30 d. The EHL@AgNPRs exhibited remarkable selectivity towards S2- with a minimum detection limit of 41.3 nM. The conjugation of lignin with AgNPRs offers a highly promising approach for the rapid detection of S2- in natural aquatic environments and for the valorization of lignin.

Spatial and temporal patterns of above- and below- ground biomass over the Tibet Plateau grasslands and their sensitivity to climate change.

The sensitivity of grassland above- (AGB, gC m-2) and below-ground biomass (BGB, gC m-2) to climate has been shown to be significant on the Tibetan Plateau, however, the spatial patterns and sensitivity of biomass with altitudinal change needs to be quantitated. In this study, large data sets of AGB and BGB during the peak growth season, and the corresponding geographical and climate conditions in the grasslands of the Tibetan Plateau between 2001 and 2020 were analyzed, and modelled using a Cubist regression trees algorithm. The mean values for AGB and BGB were 61.3 and 1304.3 gC m-2, respectively, for the whole region over the two decades. There was a significant change in spatial AGB of 64.8 % on the Plateau (P < 0.05, with areas where AGB increased being twice as large as areas where AGB decreased), while BGB did not change significantly in majority the of the region (≥ 90.1 %, P > 0.05). In general, the areas where AGB showed positive partial correlations with precipitation were larger than the areas where AGB had positive correlations with temperature (P < 0.05). However, these trends varied depending on the climatic conditions: in the wetter regions, temperature had a greater effect on the size of the areas with positive AGB responses than precipitation (P < 0.05), while precipitation had a greater effect on the size of areas with positive BGB changes than temperature (P < 0.05). In the drier areas, however, precipitation affected the AGB response significantly compared to temperature (P < 0.05), while temperature influenced the BGB response greater than precipitation (P < 0.05). The response and sensitivity of grassland biomass to temperature and precipitation varied according to the altitude of the Plateau: the response and sensitivity were stronger and more sensitive at medium altitudes, and weak at the higher or lower altitudes. Likely, this phenomenon was resulted from the natural selection of plants to maintain the efficient use of resources during un-favourable and stressed conditions for maximum plant development and growth. These findings will help assess the ecological consequences of global climate change for the grasslands of the Tibetan Plateau, particularly in those regions with highly variable altitudes.

Shaping immune landscape of colorectal cancer by cholesterol metabolites.

Cancer immunotherapies have achieved unprecedented success in clinic, but they remain largely ineffective in some major types of cancer, such as colorectal cancer with microsatellite stability (MSS CRC). It is therefore important to study tumor microenvironment of resistant cancers for developing new intervention strategies. In this study, we identify a metabolic cue that determines the unique immune landscape of MSS CRC. Through secretion of distal cholesterol precursors, which directly activate RORγt, MSS CRC cells can polarize T cells toward Th17 cells that have well-characterized pro-tumor functions in colorectal cancer. Analysis of large human cancer cohorts revealed an asynchronous pattern of the cholesterol biosynthesis in MSS CRC, which is responsible for the abnormal accumulation of distal cholesterol precursors. Inhibiting the cholesterol biosynthesis enzyme Cyp51, by pharmacological or genetic interventions, reduced the levels of intratumoral distal cholesterol precursors and suppressed tumor progression through a Th17-modulation mechanism in preclinical MSS CRC models. Our study therefore reveals a novel mechanism of cancer-immune interaction and an intervention strategy for the difficult-to-treat MSS CRC.

LRP12 is an endogenous transmembrane inactivator of α4 integrins.

Dynamic regulation of integrin activation and inactivation is critical for precisely controlled cell adhesion and migration in physiological and pathological processes. The molecular basis for integrin activation has been intensively studied; however, the understanding of integrin inactivation is still limited. Here, we identify LRP12 as an endogenous transmembrane inhibitor for α4 integrin activation. The LRP12 cytoplasmic domain directly binds to the integrin α4 cytoplasmic tail and inhibits talin binding to the ß subunit, thus keeping integrin inactive. In migrating cells, LRP12-α4 interaction induces nascent adhesion (NA) turnover at the leading-edge protrusion. Knockdown of LRP12 leads to increased NAs and enhanced cell migration. Consistently, LRP12-deficient T cells show an enhanced homing capability in mice and lead to aggravated chronic colitis in a T cell-transfer colitis model. Altogether, LRP12 is a transmembrane inactivator for integrins that inhibits α4 integrin activation and controls cell migration by maintaining balanced NA dynamics.

Controllable synthesis of carbon coated iron-based composite nanoparticles.

Different types of carbon coated iron (Fe) based nanoparticles were controllably synthesized by detonation decomposition of urea nitrate metal complex explosive precursors. In order to control the detonation synthesis of the Fe-based composite nanoparticles, the main components of the urea nitrate metal complex explosive precursors were optimally designed. The components, morphologies, structures and various Fe-based phase composition of the as-obtained composite nanoparticles were further investigated by x-ray diffraction, transmission electron microscopy, electron diffraction spectroscopy and energy dispersive x-ray spectroscopy. The results show that discrete, spherical, uniformly sized (~55 nm) iron oxide nanocrystals and core-shell structural composite nanoparticles (graphite coated Fe) were effectively synthesized by adjusting the densities, oxygen balances and mole ratios of C:Fe of these compound explosive precursors. The formation mechanism of these carbon coated Fe-based composites during the instantaneous detonation reaction process is further discussed.

Practical guide to using Kendall's τ in the context of forecasting critical transitions.

Recent studies demonstrate that trends in indicators extracted from measured time series can indicate an approach of an impending transition. Kendall's τ coefficient is often used to study the trend of statistics related to the critical slowing down phenomenon and other methods to forecast critical transitions. Because statistics are estimated from time series, the values of Kendall's τ are affected by parameters such as window size, sample rate and length of the time series, resulting in challenges and uncertainties in interpreting results. In this study, we examine the effects of different parameters on the distribution of the trend obtained from Kendall's τ, and provide insights into how to choose these parameters. We also suggest the use of the non-parametric Mann-Kendall test to evaluate the significance of a Kendall's τ value. The non-parametric test is computationally much faster compared with the traditional parametric auto-regressive, moving-average model test.

Effect of Long-Term Fertilization on Aggregate Size Distribution and Nutrient Accumulation in Aeolian Sandy Soil.

Soil aggregates are the material basis of soil structure and important carriers of nutrients. Long-term application of organic and inorganic fertilizers can affect the composition of soil aggregates to varying degrees, which in turn affects the distribution and storage of soil nutrients. We report the results of a 15-year long-term field-based test of aeolian sandy soil and used the wet sieve method to analyze the stability of water-stable aggregates, as well as the distribution characteristics of nutrients in different particle size aggregates. Our results show that long-term application of organic fertilizer (M3) and combined organic−inorganic treatments (NPK1-M1, NPK1-M2, and NPK1-M3) help to increase the amount of organic carbon, inorganic carbon, and cation exchange in the macro-aggregates, and the improvement rates are 92−103%, 8−28%, and 74−85%, respectively. The organic content of the fertilizers also promotes the formation of macro-aggregates, and the stability of aggregates increase from 0.24 to 0.45. In contrast, the application of inorganic fertilizers (NPK1, NPK2, and NPK3) has no marked effect on the formation and stability of macro-aggregates; the application of inorganic fertilizers can merely maintain the organic carbon content of the soil. Correlation analysis shows that the application of organic fertilizers and chemical (inorganic) fertilizers containing phosphorus and potassium can markedly increase the content and reserves of available phosphorus and potassium across all aggregate sizes, and there is a significant positive correlation between these parameters and the amount of applied fertilizer (p < 0.05). Aggregates of various sizes in aeolian sandy soils in arid areas have the potential for greater nutrient storage. Therefore, organic fertilizers can be used in the agricultural production process to improve soil structure and fertility.

Machine learning of pair-contact process with diffusion.

The pair-contact process with diffusion (PCPD), a generalized model of the ordinary pair-contact process (PCP) without diffusion, exhibits a continuous absorbing phase transition. Unlike the PCP, whose nature of phase transition is clearly classified into the directed percolation (DP) universality class, the model of PCPD has been controversially discussed since its infancy. To our best knowledge, there is so far no consensus on whether the phase transition of the PCPD falls into the unknown university classes or else conveys a new kind of non-equilibrium phase transition. In this paper, both unsupervised and supervised learning are employed to study the PCPD with scrutiny. Firstly, two unsupervised learning methods, principal component analysis (PCA) and autoencoder, are taken. Our results show that both methods can cluster the original configurations of the model and provide reasonable estimates of thresholds. Therefore, no matter whether the non-equilibrium lattice model is a random process of unitary (for instance the DP) or binary (for instance the PCP), or whether it contains the diffusion motion of particles, unsupervised learning can capture the essential, hidden information. Beyond that, supervised learning is also applied to learning the PCPD at different diffusion rates. We proposed a more accurate numerical method to determine the spatial correlation exponent [Formula: see text], which, to a large degree, avoids the uncertainty of data collapses through naked eyes.

Transfer learning of phase transitions in percolation and directed percolation.

The latest advances of statistical physics have shown remarkable performance of machine learning in identifying phase transitions. In this paper, we apply domain adversarial neural network (DANN) based on transfer learning to studying nonequilibrium and equilibrium phase transition models, which are percolation model and directed percolation (DP) model, respectively. With the DANN, only a small fraction of input configurations (two-dimensional images) needs to be labeled, which is automatically chosen, to capture the critical point. To learn the DP model, the method is refined by an iterative procedure in determining the critical point, which is a prerequisite for the data collapse in calculating the critical exponent ν_{⊥}. We then apply the DANN to a two-dimensional site percolation with configurations filtered to include only the largest cluster which may contain the information related to the order parameter. The DANN learning of both models yields reliable results which are comparable to the ones from Monte Carlo simulations. Our study also shows that the DANN can achieve quite high accuracy at much lower cost, compared to the supervised learning.