Maximum Correlation Research Articles

Incidence of Insect Pests and Its Natural Enemies of Cabbage (Brassica oleracea var. capitata) in Relation to Weather Parameters under Field Conditions

The study was carried out in the field at the Horticultural Polyhouse, College of Agriculture, Rajendranagar, during the rabi season of 2022–2023. In the field, aphid infestation peaked in the last week of November (48th SMW), while lepidopteran pests (tobacco caterpillar, DBM, and leaf webber) peaked in January (3rd, 2nd, and 1st SMW, respectively). Coccinellids, syrphids, and spiders had average populations of 0.80, 0.21, and 0.03 correspondingly. Aphids were shown to have a non-significant link with weather factors, showing negative correlations with maximum temperature, sunshine, wind, and rainfall and positive correlations with minimum temperature and RH I & II. While leaf webber and DBM showed a non-significant negative connection with sunshine and wind speed, they displayed a non-significant positive correlation with maximum temperature, RH I & II, and rainfall. Additionally, they showed a strong inverse relationship with evaporation. The maximum and minimum temperatures were shown to have a non-significantly positive connection with tobacco caterpillar and semilooper.

Automatic Depth Shifting by Identifying and Matching Events on Well Logs Chicheng Xu

Depth matching or depth shifting between well logs acquired from different runs or between core scans and well logs is a critical data quality control task to ensure subsequent accurate petrophysical interpretation and modeling. Conventional depth-shifting workflow heavily relies on human expertise to manually match a series of peaks and troughs between log curves, which is often subjective, error-prone, and cumbersome. Therefore, it is necessary to establish an automatic depth-shifting workflow to perform this routine yet important task accurately in a consistent and efficient manner. We implemented an automatic workflow to emulate human expertise to identify important “events” such as peaks, troughs, and bed boundaries on log curves and then intelligently match the identified series of events between log curves with local maximum correlation criteria to generate a depth shift table. We applied the automatic workflow in a field case to shift the well log and core gamma ray and delivered a depth shift table comparable to manual depth matching. The final shifted log achieved a significantly enhanced correlation with the reference log. The events identifying and matching method presents a white-box solution that still follows the conventional petrophysical wisdom and allows more user interaction to fine-tune the results. The users have full control of the parameters for optimal results.

A low-profile high-isolation endfire MIMO antenna based on spoof surface plasmon polaritons for X-band applications

This work proposes a three-port, low-profile, coplanar endfire multiple-input-multiple-output (MIMO) antenna based on spoof surface plasmon polaritons (SSPPs). The proposed antenna comprises a coplanar waveguide (CPW)-fed SSPPs transmission line (TL). A transition to transform the CPW quasi-TEM mode to the SSPPs mode. The radiator section with asymmetric discontinuities converts the bounded SSPPs waves into radiating waves. The proposed MIMO antenna exploits the inherent feature of high isolation in the SSPPs TL due to its strong field confinement competency. Wider beamwidths for ample azimuth and elevation coverage with a stable pattern in the endfire direction are achieved in the 9.52–11.23 GHz operating band. In the operating band, the MIMO antenna has an average realized gain of 6 dBi. Mainly, it has high isolation below -35 dB (measured) in the operating range, a maximum envelope correlation coefficient (ECC) of 1.5 × 10−3, and a total active reflection coefficient (TARC) value below 0.3. The measured results are closely matched with the simulated ones in terms of S-parameters, radiation pattern, ECC, TARC, and CCL.

Algorithm for Correlation Diagnosis in Multivariate Process Quality Based on the Optimal Typical Correlated Component Pair Group

Correlation diagnosis in multivariate process quality management is an important and challenging issue. In this paper, a new approach based on the optimal typical correlated component pair group (OTCCPG) is proposed. Firstly, the theorem of correlation decomposition is proved to decompose the correlation of all the quality components as serial correlations of component pairs, and then according to the transitivity of correlations of component pairs, the decomposition result is represented by a correlation set of typical correlated component pairs. Finally, an algorithm for OTCCPG based on the maximum correlation spanning tree (MCST) is proposed, and T2 control charts to monitor the correlations of component pairs in OTCCPG are established to form the correlation diagnostic system. Theoretical analysis and practice prove that the proposed method could reduce the space complexity of the diagnostic system greatly.

3D suitability evaluation of urban underground space using a variable weight method and considering ground restrictions

The evaluation of urban underground space (UUS) suitability involves multiple indicators. Assigning weight to these indicators is crucial for accurate assessment. This paper presents a method for spatially variable weight assignment of indicators using the order relation analysis method (G1-method), the entropy weight method, an improved grey relational analysis (GRA) and a set of spatial weight adjustment coefficients. First, the subjective and objective weights of indicators for engineering geological and hydrogeological conditions were determined by the G1-method and entropy weight method, respectively, and their combined weights were then obtained using the principle of minimum discriminatory information. This study highlighted the impact of surface restrictions, such as buildings, on UUS, and the degree of the influence of these buildings gradually decreased with the increase in depth of the rock and soil mass in UUS, which resulted in changes in weights of indicators with depth. To address this issue, a coefficient was defined as the standardized value of the ratio of additional stress applied by restrictions to the self-weight stress of soil at the same depth to modify the combined weights so that all weights of indicators could vary in space. Finally, an improved GRA was used to determine the suitability level of each evaluation cell using the maximum correlation criterion. This method was applied to the 3D suitability evaluation of UUS in Sanlong Bay, Foshan City, Guangdong Province, China, including 16 evaluation indexes. This study comprehensively considered the influence of multiple factors, thereby providing reference for evaluating the suitability of UUS in big cities.

H-ensemble: An Information Theoretic Approach to Reliable Few-Shot Multi-Source-Free Transfer

Multi-source transfer learning is an effective solution to data scarcity by utilizing multiple source tasks for the learning of the target task. However, access to source data and model details is limited in the era of commercial models, giving rise to the setting of multi-source-free (MSF) transfer learning that aims to leverage source domain knowledge without such access. As a newly defined problem paradigm, MSF transfer learning remains largely underexplored and not clearly formulated. In this work, we adopt an information theoretic perspective on it and propose a framework named H-ensemble, which dynamically learns the optimal linear combination, or ensemble, of source models for the target task, using a generalization of maximal correlation regression. The ensemble weights are optimized by maximizing an information theoretic metric for transferability. Compared to previous works, H-ensemble is characterized by: 1) its adaptability to a novel and realistic MSF setting for few-shot target tasks, 2) theoretical reliability, 3) a lightweight structure easy to interpret and adapt. Our method is empirically validated by ablation studies, along with extensive comparative analysis with other task ensemble and transfer learning methods. We show that the H-ensemble can successfully learn the optimal task ensemble, as well as outperform prior arts.

Morphological characterization of Maize (Zea mays.) utilising the stage-wise structural and architectural perspective from temporal fully-polarimetric SAR

The morphological shape and structure of the crop vary with phenological stages. Model and eigen based decomposition model parameters extracted from the Radarsat-2 data and the trend with respect to ground truth crop phenology were analysed. Sensitive parameters were devised through stepwise approach under 7 combinations of polarimetric variables of increasing complexity were assessed. Compared under the three machine learning algorithms (ANN, RF and SVM) where ANN rendered the maximum correlation with 0.92 with a MAE of 4 days which was implemented on a large parcel of maize mask in the study area. SVM performed poorly with highly overlapping parameters such as backscatter but performed well (r = 0.85). For assessing the crop biophysical parameters, the three algorithms were evaluated and sensitivity analysis for statistically significant polarimetric variables for biophysical parameters was performed. The assessment was performed on Multi-Layer Perception (MLP) neural network. The networks were trained with algorithms and hidden layer nodes until the MAE achieved permissible limits. Plant height could be estimated more profoundly with an r = 0.8 with a considerably good MAE of 24.9 cm but other parameters (WB, DB and LAI) were estimated in moderate correlation of 0.6–0.65 where the MAE of WB, DB and LAI were found to be 1317gm−2, 553 gm−2 and 0.78 respectively. This is the first step towards understanding the complex scattering mechanisms in Indian maize scenario assessing the growth parameters from polarimetric data. Thus, the analytical findings brought out possess the potential to serve as the reference for the future research initiatives.

A new approach of multi-dimensional correlation as a separability measure of multiple outliers in GNSS applications

Abstract Detecting and identifying outliers/failures in GNSS measurements has garnered significant attention among researchers aiming to enhance the quality of GNSS positioning and navigation. This study delves into the analysis of the separability of multiple outliers when single, double, and triple outliers occur in single-point positioning (SPP) measurements. To achieve this, a novel method includes introducing a multi-dimensional correlation coefficient among test statistics. This coefficient functions as a measure of outliers separability and, in turn, assesses the possible impact of outliers on other measurements. This multi-dimensional correlation approach is based on a nested correlation ( ρ nested θ , φ ${\rho }_{\text{nested}}^{\theta ,\varphi }$ ) that explains the variations in test statistic values with and without common measurements in two pairs/combinations. The performance of the ρ nested θ , φ ${\rho }_{\text{nested}}^{\theta ,\varphi }$ is then compared with other two existing methods of multi-dimensional correlation namely the maximum ( ρ max θ , φ ${\rho }_{\mathrm{max}}^{\theta ,\varphi }$ ) and global ( ρ Global θ , φ ${\rho }_{\text{Global}}^{\theta ,\varphi }$ ) correlation. The results show that under the presence of two outliers and with and without common measurements in two pairs, the ρ nested θ ${\rho }_{\text{nested}}^{\theta }$ outperforms the, ρ max θ ${\rho }_{\mathrm{max}}^{\theta }$ and ρ Global θ ${\rho }_{\text{Global}}^{\theta }$ exhibiting a determination coefficient (R 2) value of approximately 0.95 and 0.62 respectively. The results furthermore reveal that for three outliers and with one, two, and noncommon measurements intersecting between two combinations, the values of R 2 are 0.62, 0.96, and 0.34. respectively. This means that the ρ nested θ , φ ${\rho }_{\text{nested}}^{\theta ,\varphi }$ can explain the variations in outlier test statistic values particularly in the case that common measurements appear in two pairs/combinations.

Frequency and spectral analysis of pulsing aurora according to the data of the imaging photometer at the Verkhnetulomsky observatory

A Pulsating Aurora Imaging Photometers System (PAIPS) is being created using observatories of the Polar Geophysical Institute on the Kola Peninsula to study the structure of pulsating auroras (PsA) in the near ultraviolet range (300—400 nm). We presented a description of two PsA cases measured by the PAIPS system together with all-sky cameras. For the first one, a frequency analysis of the signal is given and it is shown that for an hour and a half, pulsations occur at a quasi-constant frequency of about 1 Hz. For the second one, measurements of the intensity ratio in the 337 and 391 nm wavelengths are presented. An analysis of simultaneous observations of optical cameras in the field of view of the spectrometer shows that the emission in the “on” phase is “greener”, i.e., corresponds to less energetic precipitating electrons. Stereometric observations with a camera in observatory Lovozero gives the maximum correlation coefficient for the time series of cameras at altitudes of 145—150 km and ~92 km.

A new automatic geo-electric self-potential imaging technique for diverse sustainable development scenarios

This study introduces a rapid and efficient inversion algorithm designed for the interpretation of self-potential responses originating from mineralized and ore sources and hydrothermal activity, specifically addressing spherical, vertical, and horizontal cylindrical structures. The algorithm leverages local wavenumber and correlation imaging techniques to enhance accuracy in modeling. The correlation factor (Cf value) is crucial in this approach, calculated as the correlation between the local wavenumber of the measured self-potential field and that of the computed field. The algorithm identifies the maximum correlation Cf value (CF-max) as indicative of the optimal true model parameters. To validate the proposed algorithm, it was applied to three theoretical examples—one with contamination from regional background and another with multiple sources with and without different types of noises (random Gaussian and white Gaussian noises). Additionally, the approach was tested on three distinct real field cases related to mining, ore investigation and hydrothermal activity in India, Germany and USA. Through a comprehensive analysis of results from theoretical and real-world scenarios, including comparisons with different available data and literature information, the study concludes that the method is effective, applicable to multiple sources, accurate, and does not necessitate prior knowledge of the source shape. This algorithm presents a promising advancement in the field of self-potential interpretation for mineral exploration and geothermal exploration.

Research on SPAD Estimation Model for Spring Wheat Booting Stage Based on Hyperspectral Analysis.

With the rapid progression of agricultural informatization technology, the methodologies of crop monitoring based on spectral technology are constantly upgraded. In order to carry out the efficient, precise and nondestructive detection of relative chlorophyll (SPAD) during the booting stage, we acquired hyperspectral reflectance data about spring wheat vertical distribution and adopted the fractional-order differential to transform the raw spectral data. After that, based on correlation analysis, fractional differential spectra and fractional differential spectral indices with strong correlation with SPAD were screened and fused. Then, the least-squares support vector machine (LSSSVM) and the least-squares support vector machine (SMA-LSSSVM) optimized on the slime mold algorithm were applied to construct the estimation models of SPAD, and the model accuracy was assessed to screen the optimal estimation models. The results showed that the 0.4 order fractional-order differential spectra had the highest correlation with SPAD, which was 9.3% higher than the maximum correlation coefficient of the original spectra; the constructed two-band differential spectral indices were more sensitive to SPAD than the single differential spectra, in which the correlation reached the highest level of 0.724. The SMA-LSSSVM model constructed based on the two-band fractional-order differential spectral indices was better than the single differential spectra and the integration of both, which realized the assessment of wheat SPAD.

Research and Application of Power Grid Data Anomaly Tracing Method Based on Time Series Correlation

Abstract Power grid data is a “barometer” that reflects the operational trend of the power grid system, so conducting a traceability study on abnormal data generated by the power grid system is crucial to maintaining the stable operation of the power grid system and preventing further malfunctions. This article proposes a maximum time series correlation ring tracing algorithm based on time series correlation: firstly, for a large amount of time series data collected from measurement points in the power grid system, a correlation coefficient matrix is calculated, and then a time series correlation graph is constructed using graph theory knowledge. The Kruskal algorithm is used to search for the longest spanning tree in the time series correlation graph. Finally, based on the spanning tree, the breadth-first search (BFS) algorithm is further used to obtain the maximum time series correlation ring[1]. By using the time series correlation of each node within the ring and the physical topological relationship between nodes, the abnormal data can be traced back. Distinguish whether the generation of abnormal data is due to a single point failure or system failure. Most of the existing fault tracing techniques are based on machine learning algorithms, which are difficult to widely apply in high-dimensional time series data due to their high complexity.Through experiments on real power grid data, the results verify the effectiveness of this method in anomaly tracing of high-dimensional time series data. Through comparative experiments, this method is superior to machine learning model algorithms in terms of efficiency and stability. At the same time, this method does not require additional calculations of the statistical distribution of the samples to be tested, thus greatly saving computational costs.

Джаз в советском кино. Опыт ретроспективного анализа

The article is devoted to representation of domestic jazz in the Soviet feature silent and sound cinema in the period from 1926 to 1990. The author notes that jazz was becoming part of the plot of documentaries and feature films since its emergence in the USSR, but the accents in its image changed, and the genre nature of the movies evolved. There was a traceable shift from musical comedy to satirical pamphlet, from lyrical comedy to melodrama. Based on several dozens of examples of films where jazz sounds, the article shows the relationship between the ideological course of the Soviet state, the perception of the jazz performance specifics at different historical milestones, and the nature of its reflection on the screen. Thus, in the 1930s, jazz was synonymous with popular music and had a clear shade of humour and entertainment; in the late 1940s, it symbolized the ideological opponents of the country; and after the Khrushchev Thaw, it was considered an art, a kind of intellectual activity. In addition to the periodization of jazz cinema, the article provides a typology of films about jazz. The author presents a classification based on a set of qualities, such as maximum or minimum correlation of the plot of the film with jazz, participation of jazz musicians in it, and presence of a jazz soundtrack.

Speech-driven head motion generation from waveforms

Head motion generation task for speech-driven virtual agent animation is commonly explored with handcrafted audio features, such as MFCCs as input features, plus additional features, such as energy and F0 in the literature. In this paper, we study the direct use of speech waveform to generate head motion. We claim that creating a task-specific feature from waveform to generate head motion leads to better performance than using standard acoustic features to generate head motion overall. At the same time, we completely abandon the handcrafted feature extraction process, leading to more effectiveness. However, the difficulty of creating a task-specific feature from waveform is their staggering quantity of irrelevant information, implicating potential cumbrance for neural network training. Thus, we apply a canonical-correlation-constrained autoencoder (CCCAE), where we are able to compress the high-dimensional waveform into a low-dimensional embedded feature, with the minimal error in reconstruction, and sustain the relevant information with the maximal cannonical correlation to head motion. We extend our previous research by including more speakers in our dataset and also adapt with a recurrent neural network, to show the feasibility of our proposed feature. Through comparisons between different acoustic features, our proposed feature, WavCCCAE, shows at least a 20% improvement in the correlation from the waveform, and outperforms the popular acoustic feature, MFCC, by at least 5% respectively for all speakers. Through the comparison in the feedforward neural network regression (FNN-regression) system, the WavCCCAE-based system shows comparable performance in objective evaluation. In long short-term memory (LSTM) experiments, LSTM-models improve the overall performance in normalised mean square error (NMSE) and CCA metrics, and adapt the WavCCCAEfeature better, which makes the proposed LSTM-regression system outperform the MFCC-based system. We also re-design the subjective evaluation, and the subjective results show the animations generated by models where WavCCCAEwas chosen to be better than the other models by the participants of MUSHRA test.

Construction and Analysis of a New Resting-State Whole-Brain Network Model.

Mathematical modeling and computer simulation are important methods for understanding complex neural systems. The whole-brain network model can help people understand the neurophysiological mechanisms of brain cognition and functional diseases of the brain. In this study, we constructed a resting-state whole-brain network model (WBNM) by using the Wendling neural mass model as the node and a real structural connectivity matrix as the edge of the network. By analyzing the correlation between the simulated functional connectivity matrix in the resting state and the empirical functional connectivity matrix, an optimal global coupling coefficient was obtained. Then, the waveforms and spectra of simulated EEG signals and four commonly used measures from graph theory and small-world network properties of simulated brain networks under different thresholds were analyzed. The results showed that the correlation coefficient of the functional connectivity matrix of the simulated WBNM and empirical brain networks could reach a maximum value of 0.676 when the global coupling coefficient was set to 20.3. The simulated EEG signals showed rich waveform and frequency-band characteristics. The commonly used graph-theoretical measures and small-world properties of the constructed WBNM were similar to those of empirical brain networks. When the threshold was set to 0.22, the maximum correlation between the simulated WBNM and empirical brain networks was 0.709. The constructed resting-state WBNM is similar to a real brain network to a certain extent and can be used to study the neurophysiological mechanisms of complex brain networks.

Radiomics-Based Prediction Model for Outcome of Radioembolization in Metastatic Colorectal Cancer.

To evaluate the benefit of a contrast-enhanced computed tomography (CT) radiomics-based model for predicting response and survival in patients with colorectal liver metastases treated with transarterial Yttrium-90 radioembolization (TARE). Fifty-one patients who underwent TARE were included in this single-center retrospective study. Response to treatment was assessed using the Response Evaluation Criteria in Solid Tumors (RECIST 1.1) at 3-month follow-up. Patients were stratified as responders (complete/partial response and stable disease, n = 24) or non-responders (progressive disease, n = 27). Radiomic features (RF) were extracted from pre-TARE CT after segmentation of the liver tumor volume. A model was built based on a radiomic signature consisting of reliable RFs that allowed classification of response using multivariate logistic regression. Patients were assigned to high- or low-risk groups for disease progression after TARE according to a cutoff defined in the model. Kaplan-Meier analysis was performed to analyze survival between high- and low-risk groups. Two independent RF [Energy, Maximal Correlation Coefficient (MCC)], reflecting tumor heterogeneity, discriminated well between responders and non-responders. In particular, patients with higher magnitude of voxel values in an image (Energy), and texture complexity (MCC), were more likely to fail TARE. For predicting treatment response, the area under the receiver operating characteristic curve of the radiomics-based model was 0.75 (95% CI 0.48-1). The high-risk group had a shorter overall survival than the low-risk group (3.4 vs. 6.4months, p < 0.001). Our CT radiomics model may predict the response and survival outcome by quantifying tumor heterogeneity in patients treated with TARE for colorectal liver metastases.

Relativistic electron ([formula omitted]) flux driving parameters over geostationary orbit using 30 years of data

The relativistic electron population at MeV energy in the Earth’s outer radiation belt at geostationary (GOE) orbit largely varies in association with solar wind disturbances. Although more researches have been done to understand the influence of different mechanisms that cause the variations in the relativistic electron flux in the region, but still the underlying physics is often complex and unclear. The present study uses relativistic electron (E>2MeV) fluxes data of Los Alamos National Laboratory (LANL) geosynchronous satellites to understand the variations and its relationship with the driving parameters. The study examines the relativistic electron fluxes variations associated with the interplanetary magnetic field (IMF) and solar wind parameters using daily averaged data from 01 January 1990 to 31 December 2018. The results indicated that the information transfer from solar wind and IMF parameters into relativistic electrons flux is dependent on the kind of driving parameter. The result identifies the rank of driving parameters based on their maximum values in the cross correlation and information transfer analysis. Thus, AE index and vsw are the dominant driving parameters for the dynamics of relativistic electrons flux at GEO orbit. The peak values of coefficient (correlation coefficients, mutual information and entropy transfer) are observed at similar lag times (τpeak). For instance, in the case of AE and Je, we have observed the maximum coefficient (associated with lag time) as follows; maximum correlation coefficient; CCmax=0.846(5days), maximum mutual information; MImax=0.803(5days), maximum transfer entropy; TEmax=0.892(5days). Such dependency analysis can help to select the model parameters that influence the relativistic electrons fluxes and improve the existed models of the dynamical system.

Attitude and heading measurement based on adaptive complementary Kalman filter for PS/MIMU integrated system.

The bionic polarization sensor (PS)/MEMS inertial measurement unit (MIMU) integrated system can provide reliable attitude and heading information for unmanned vehicles in the case of GNSS rejection. However, the existing measurement methods have poor adaptability to inclining, sheltering, and other harsh environments, and do not make full use of the complementary characteristics of the gyroscopes, accelerometers, and PS, which seriously affects the system performance. Therefore, this paper proposes an attitude and heading measurement method based on an adaptive complementary Kalman filter (ACKF), which corrects the gyroscopes according to the gravity measured by the accelerometers to improve the attitude accuracy and fuses the IMU heading and tilt-compensated polarization heading by Kalman optimal estimation. On this basis, the maximum correlation entropy of the measured gravity and the theoretical gravity is used to construct an adaptive factor to realize the adaptive complementary of the gyroscopes and the accelerometers. Finally, the effectiveness of the method is verified by the outdoor rotation test without occlusion and the vehicle test with occlusion. Compared with the traditional Kalman filter, the pitch, roll, and heading RMSE of the vehicle test are reduced by 89.3%, 93.2% and, 9.6% respectively, which verifies the great advantages.

Fault diagnosis of tooth surface spalling based on variational mode decomposition and maximum correlation kurtosis method

As the early fault features of tooth surface spalling are very weak and difficult to extract because of random noise and other types of signal interference, a method that combines maximum correlation kurtosis uncoiling and variational mode decomposition is proposed herein. First, a series of modes are obtained by variational mode decomposition, and the kurtosis criterion is applied to select the modes containing rich fault information for reconstruction and noise reduction. Second, the maximum correlation kurtosis deconvolution method is used to enhance the selected signals. Finally, the fault features are extracted by envelope demodulation of the reconstructed signal. The effectiveness of the proposed method is verified by analysis, and the different frequency components of the vibration signals of tooth surface spalling faults are shown to be separated accurately.

A Keratin 12 Expression-Based Analysis of Stem-Precursor Cells and Differentiation in the Limbal-Corneal Epithelium Using Single-Cell RNA-Seq Data.

The corneal epithelium (CE) is spread between two domains, the outer vascularized limbus and the avascular cornea proper. Epithelial cells undergo constant migration from the limbus to the vision-critical central cornea. Coordinated with this migration, the cells undergo differentiation changes where a pool of unique stem/precursor cells at the limbus yields the mature cells that reach the corneal center. Differentiation is heralded by the expression of the corneal-specific Krt12. Processing data acquired by scRNA-Seq showed that the increase in Krt12 expression occurs in four distinct steps within the limbus, plus a single continuous increase in the cornea. Differential gene analysis demonstrated that these domains reflect discreet stages of CE differentiation and yielded extensive information of the genes undergoing down- or upregulation in the sequential transition from less to more differentiate conditions. The approach allowed the identification of multiple gene cohorts, including (a) the genes which have maximal expression in the most primitive, Krt12-negative cell cohort, which is likely to include the stem/precursor cells; (b) the sets of genes that undergo continuous increase or decrease along the whole differentiation path; and (c) the genes showing maximal positive or negative correlation with the changes in Krt12.