Log-ratio Transformation Research Articles

Forecasting age distribution of life-table death counts via α-transformation

We introduce a compositional power transformation, known as an α-transformation, to model and forecast a time series of life-table death counts, possibly with zero counts observed at older ages. As a generalisation of the isometric log-ratio transformation (i.e. α = 0 ), the α transformation relies on the tuning parameter α, which can be determined in a data-driven manner. Using the Australian age-specific period life-table death counts from 1921 to 2020, the α transformation can produce more accurate short-term point and interval forecasts than the log-ratio transformation. The improved forecast accuracy of life-table death counts is of great importance to demographers and government planners for estimating survival probabilities and life expectancy and actuaries for determining annuity prices and reserves for various initial ages and maturity terms.

The choice of 16S rRNA gene sequence analysis impacted characterization of highly variable surface microbiota in dairy processing environments.

Accurate knowledge of the microbiota collected from surfaces in food processing environments is important for food quality and safety. This study assessed discrepancies in taxonomic composition and alpha and beta diversity values generated from eight different bioinformatic workflows for the analysis of 16S rRNA gene sequences extracted from the microbiota collected from surfaces in dairy processing environments. We found that the microbiota collected from environmental surfaces varied widely in density (0-9.09 log10 CFU/cm2) and Shannon alpha diversity (0.01-3.40). Consequently, depending on the sequence analysis method used, characterization of low-abundance genera (i.e., below 1% relative abundance) and the number of genera identified (114-173 genera) varied considerably. Some low-abundance genera, including Listeria, varied between the amplicon sequence variant (ASV) and operational taxonomic unit (OTU) methods. Centered log-ratio transformation inflated alpha and beta diversity values compared to rarefaction. Furthermore, the ASV method also inflated alpha and beta diversity values compared to the OTU method (P < 0.05). Therefore, for sparse, uneven, low-density data sets, the OTU method and rarefaction are better for taxonomic and ecological characterization of surface microbiota.IMPORTANCECulture-dependent environmental monitoring programs are used by the food industry to identify foodborne pathogens and spoilage biota on surfaces in food processing environments. The use of culture-independent 16S rRNA amplicon sequencing to characterize this surface microbiota has been proposed as a tool to enhance environmental monitoring. However, there is no consensus on the most suitable bioinformatic analyses to accurately capture the diverse levels and types of bacteria on surfaces in food processing environments. Here, we quantify the impact of different bioinformatic analyses on the results and interpretation of 16S rRNA amplicon sequences collected from three cultured dairy facilities in New York State. This study provides guidance for the selection of appropriate 16S rRNA analysis procedures for studying environmental microbiota in dairy processing environments.

A rapid method for measuring the rock brittleness index: Rapid characterization of rock brittleness based on LIBS technology

The brittle index is a crucial indicator in the assessment of rockbursts. The mineral brittle index (MBI) is widely utilized due to its simplicity and accessibility. However, the lengthy and inefficient mineral composition testing cycle presents a significant challenge. A novel approach using laser-induced breakdown spectroscopy (LIBS) to rapidly convert spectral elements into minerals and measure the rock brittleness index was introduced in this paper. The laser spectra of metamorphic sandstone and granite were measured by LIBS, and some rock elements were tested by X-ray fluorescence (XRF) to construct a spectral-elemental model. Furthermore, the mineral composition of the rocks was determined by X-ray diffraction (XRD). Eight major elements, Si, Al, K, Ca, Na, Fe, Mg, and Ti, were selected as independent variables. Element-mineral correlation analysis was performed using centered log-ratio transformation (CLR) processing. A random forest regression (RF-R) element-mineral transformation model was established for rapid conversion of spectral-element-mineral brittle index. Finally, mechanical testing of rock embrittlement in the DJ Tunnel was conducted to verify the efficacy of MBI in characterizing rock embrittlement in the DJ Tunnel. The results demonstrate that predicted values of mineral compositions are in good agreement with experimental values and that predicted values of brittleness indices are also in good agreement with experimental values. The rapid and effective application of LIBS in determining rock mineral composition and rock brittleness index has been realised, which is of great significance for further realising the rapid assessment of rock brittleness and rockburst prediction at engineering sites.

Geochemical vectors for targeting hydrothermal gold mineralization in the Damang area, southwestern Ghana: Constraints from petrography, multi-element geochemistry and multivariate statistical analysis

The Damang Gold Mine in southwestern Ghana sits on two kinds of mineralization: the Tarkwaian paleoplacer and a hydrothermal type found within the Tarkwaian Group's meta-sedimentary rocks. Geochemical vectors related to the hydrothermal gold mineralization in the area are not well known. Therefore, this research was conducted to identify pathfinders for hydrothermal gold mineralization in the Damang area. A total of fifty (50) rock samples were collected from three diamond drill holes focusing on mineralized zones of the holes. The mineralogy of the host rocks was established using petrographic studies of twelve (12) selected representative drill core samples. Multi-element geochemical analysis of the host rocks (n = 50) involving Inductively Coupled Plasma-Mass Spectrometry (ICP-MS) technique was conducted to quantify the trace element concentrations in the host rocks. The multi-element geochemical data was analyzed using multivariate statistics. The petrographic studies showed the abundance of quartz, biotite, pyrite, pyrrhotite, sericite, and minor amounts of feldspars (plagioclase and orthoclase). Al, Mg, Fe, and Ca dominated the geochemical data. The highest Au concentrations were found in Birimian quartz veins and the intrusive lithologies compared to the other lithologies in the region. Most of the data failed the Shapiro-Wilk and Kolmogorov-Smirnov tests. The deviations from normality were shown graphically by Q-Q plots. The centered log-ratio transformation was used to correct this, and the data tested again for normality. Spearman rank correlation showed a positive correlation between Te, Bi, U, Ba, Ag, and Au. When hierarchical cluster analysis was run, three clusters emerged with Au in the same cluster with Te, Ag, and Bi. Factor analysis served as a confirmatory test with Au in the same factor with Bi, Ag, and Te. Datamine Studio RM was used to create drillhole logs showing the anomalies of the three elements and Au with different lithologies downhole, which showed a similitude in the abnormal occurrences. Therefore, hydrothermal gold mineralization in the Damang area is related to Bi-Te-Ag-Au multi-element association.

Enhanced hybrid LSTM and SLAR modeling for in-depth analysis of temporal and spatial patterns in compositional data for environmental monitoring

A novel methodology for analyzing compositional data (CoDa) integrates long short-term memory (LSTM) networks with spatial lag autoregressive (SLAR) models to simultaneously capture temporal and spatial patterns. The proposed approach leverages LSTM’s strengths in handling temporal relationships and SLAR’s ability to capture spatial correlations, employing the centered log-ratio (CLR) transformation to manage CoDa’s characteristics, making it suitable for advanced analysis. The Adam optimizer ensures efficient and stable model training, enhancing convergence speed and precision. The combined approach involves preprocessing CoDa with CLR, initializing LSTM layers, performing forward passes through LSTM, integrating the SLAR model, generating outputs, post-processing data, calculating loss, and performing backpropagation. The methodology’s effectiveness is demonstrated through a case study on pollutant emissions data from a waste incineration plant, focusing on key pollutants such as Dioxins and Furans, Heavy Metals (Hg, Pb, Cd), Nitrogen Oxides (NO), Sulfur Dioxide (SO), Particulate Matter (PM), and Carbon Monoxide (CO). The LSTM network, configured with two hidden layers, dynamically adjusts learning rates for faster convergence. Integrating SLAR improves spatial pattern accounting, significantly enhancing predictive accuracy. Comparative analysis shows substantial performance metric improvements over standard LSTM models, with the LSTM-SLAR model reducing errors and explaining a higher proportion of data variability. In process safety and risk engineering, the proposed methodology enhances pollutant emission monitoring and control, enables proactive risk management, ensures regulatory compliance, and improves risk assessment accuracy, making it invaluable for dynamic risk assessment in environmental monitoring and is broadly applicable to fields requiring detailed temporal and spatial analysis, demonstrating robustness and versatility in handling complex CoDa.

Geochemical exploration for tantalum in coltan-rich pegmatites at Bewadze-Mankoadze area of the Kibi-Winneba Belt, southern Ghana: Constraints from exploratory data analysis

The Bewadze-Mankoadze pegmatites in the Kibi-Winneba Belt of Ghana host several columbite group minerals (WGM) and wodginite group minerals (WGM) as well as other rare and radioactive elements such as uraninite and cesium. In this study, petrographic studies of rock samples from pegmatite outcrops and statistical analysis of the major and minor elements were conducted to identify the pathfinder elements of a new tantalum deposit in the area. Ten samples were obtained from each town for whole-rock geochemistry and thin sections were prepared from some of the samples taken for petrographic analysis. The petrographic analyses showed the presence of quartz, K-feldspars, plagioclase, muscovite, spodumene, albite, tourmaline, columbite group minerals, and montebrasite, which indicate that the studied samples are granitic pegmatites. The geochemical data of the 10 samples obtained from each town showed high concentrations of Cr, Cs, Rb, Sm, and Ta. The Ta concentrations ranged from 10.5 to 773 ppm with an average value of 260 ppm. Q-Q plots showed outliers and variations from the dataset's normal distribution, which were fixed by centred log-ratio transformation and demonstrated to be normal by the Kolmogorov-Smirnov test and Shapiro-Wilk test for normality. Spearman correlation revealed that Sc, Ga, Nb, and Cs showed a moderate to strong correlation with Ta. Factor analysis indicated elemental association of Ta with Cs, Zn, Nb, MgO, Sc, Ga, and V. Three (3) multi-element relationships were discovered by hierarchical cluster analysis: (1) As, La, Hf, CaO, U, Co, Pb, Ce, Ba and Na2O, (2) V, Nb, Ta, Ga, Sc, Cr, Cu, Nb and MgO, and (3) Ni, SiO2, Cs, Zn, Sm, Rb, K2O, Y, Th, Al2O3 and Fe2O3. Hence, a comparison of the results of the multivariate statistics established Sc, Ga, Nb, and Cs as the pathfinders of tantalum in the Bewadze-Mankoadze area. Geochemical anomalies involving these elements can be observed in the south-western portion of the study area, according to single and multi-element halo mapping. It is therefore recommended that exploration activities for tantalum mineralization should focus on the south-western part of the study area, where the anomalies of the pathfinder elements are located.

토픽모델링에 기반한 R&amp;D 투자 포트폴리오 분석 및 예측

Many countries are identifying crucial technologies for national security and economic growth amid global tech competition, crafting cross-departmental strategies to pursue them. We propose a platform for analyzing and predicting R&amp;D investment portfolios based on government R&amp;D project information and investment data to provide scientific evidence for strategic budget allocation. The proposed platform consists of the following four processes. First, topic modeling and project labeling are performed using project-related text data. Second, we propose a method for calculating the importance of each label based on the centered log ratio (CLR) transformation. Third, we predict the label importance for the next year using linear regression and Holt's linear trend method. Fourth, we propose a method for constructing investment portfolio in specific fields. Real data analysis demonstrates the explanability and predictability of the proposed platform by using research project data spanning 11 years from 2012 to 2022. We believe a novelty of the proposed method can contribute to the establishment of data-driven R&amp;D investment portfolios.

Multivariate statistical analysis and bespoke deviation network modeling for geochemical anomaly detection of rare earth elements

Rare earth elements (REEs), a significant subset of critical minerals, play an indispensable role in modern society and are regarded as “industrial vitamins,” making them crucial for global sustainability. Geochemical survey data proves highly effective in delineating metallic mineral prospects. Separating geochemical anomalies associated with specific types of mineralization from the background reflecting geological processes has long been a significant subject in exploration geochemistry. The processing of high-dimensional, non-linear geochemical survey data necessitates a systematic framework to address common issues, including missing values, the closure effect, the selection of appropriate multivariate analysis methods, and anomaly detection techniques in order to detect geochemical anomalies associated with mineral occurrences. The Curnamona Province in South Australia is considered an emerging REE province with significant REE mineralization potential. In this study, we use data from this region to evaluate the performance of a novel machine learning-based framework that incorporates data pre-processing, multivariate statistical analysis, and anomaly recognition to address challenges such as missing data, noise interference, data imbalance and high non-linearity. We utilize lithogeochemical data to map potential greenfield regions of REE mineralization. The primary advantages of our framework lie in its provision of an effective random forest-based data imputation method, utilization of isometric log-ratio transformation to eliminate the closure effect, and reduction of the impact of outliers on data interpretation through robust principal component analysis. Additionally, the framework utilizes a deviation network to learn anomaly scores from complex, non-linear data under imbalanced data conditions, identifying geochemical anomalies associated with REE occurrences by leveraging prior knowledge rather than those caused by data noise or anthropogenic factors. The anomalous areas identified by this framework delineate all known REE deposits and extend to the surrounding regions. Furthermore, a close spatial coupling relationship exists between these strongly anomalous areas and the felsic granite intrusions. The comprehensive workflow for processing geochemical data proposed in this study can effectively address common challenges in the geochemical exploration of critical minerals. The identified geochemical anomalies can provide important clues for subsequent exploration.

Safety assurance and nutritional quality enhancement of Phyllospora comosa biomass using hydrothermal treatment derived ensemble machine learning models

One of the major barriers to the mass industrial utilisation of brown seaweeds as food sources stem from the food safety risks associated with their iodine and arsenic concentrations, which typically exceed regulatory limits. Hydrothermal treatments might effectively reduce the high iodine and arsenic concentration of Phyllospora comosa below the Australian maximum residual limits (iodine = 1 mg/g and arsenic = 0.00667 mg/g; dry weight) set for brown seaweeds. The experimental hydrothermal treatments dictated that the 82 °C–250 s treatment reduced the iodine concentration from 2.76 mg/g to 0.88 mg/g (68% reduction) and arsenic concentration from 0.01693 mg/g to 0.00965 mg/g (43% reduction). Machine learning models predicted that blanching at 100 °C for ∼4 minutes will reduce the arsenic concentration below its maximum residual limit. Additive log-ratio transformations showed that around 50% (dw) of the hydrothermally treated Phyllospora comosa samples were leached out during the highest treatment intensity, 82 °C–250 s. Even though, a half of the biomass is lost, hydrothermally treated Phyllospora comosa products are safer for human consumption and thus may permit the expansion of seaweed production and consumption in Australia.

Improving Likert scale big data analysis in psychometric health economics: reliability of the new compositional data approach

Bipolar psychometric scales data are widely used in psychologic healthcare. Adequate psychological profiling benefits patients and saves time and costs. Grant funding depends on the quality of psychotherapeutic measures. Bipolar Likert scales yield compositional data because any order of magnitude of agreement towards an item assertion implies a complementary order of magnitude of disagreement. Using an isometric log-ratio (ilr) transformation the bivariate information can be transformed towards the real valued interval scale yielding unbiased statistical results increasing the statistical power of the Pearson correlation significance test if the Central Limit Theorem (CLT) of statistics is satisfied. In practice, however, the applicability of the CLT depends on the number of summands (i.e., the number of items) and the variance of the data generating process (DGP) of the ilr transformed data. Via simulation we provide evidence that the ilr approach also works satisfactory if the CLT is violated. That is, the ilr approach is robust towards extremely large or infinite variances of the underlying DGP increasing the statistical power of the correlation test. The study generalizes former results pointing out the universality and reliability of the ilr approach in psychometric big data analysis affecting psychometric health economics, patient welfare, grant funding, economic decision making and profits.

Two-Part Mixed Effects Mixture Model for Zero-Inflated Longitudinal Compositional Data

Compositional data (CD) is mostly analyzed using ratios of components and log-ratio transformations to apply known multivariable statistical methods. Therefore, CD where some components equal zero represents a problem. Furthermore, when the data is measured longitudinally, and appear to come from different sub-populations, the analysis becomes highly complex. Our objective is to build a statistical model addressing structural zeros in longitudinal CD and apply it to the analysis of radiation-induced lung damage (RILD) over time. We propose a two-part mixed-effects model extended to the case where the non-zero components of the vector might come from a two-component mixture population. Maximum likelihood estimates for fixed effects and variance components were calculated by an approximate Fisher scoring procedure base on sixth-order Laplace approximation. The expectation-maximization (EM) algorithm estimates the mixture model’s probability. This model was used to analyze the radiation therapy effect on tissue change in one patient with non-small cell lung cancer (NSCLC), utilizing five CT scans over 24 months. Instead of using voxel-level data, voxels were grouped into larger subvolumes called patches. Each patch’s data is a CD vector showing proportions of dense, hazy, or normal tissue. Proposed method performed reasonably for estimation of the fixed effects, and their variability. However, the model produced biased estimates of the nuisance parameters in the model.

Area-Level Model-Based Small Area Estimation of Divergence Indexes in the Spanish Labour Force Survey

Abstract This article develops model-based predictors for area-level proportions of employed men and women by occupation sectors and for entropies and divergence indexes (DIs) within and between sex groups. Since the direct estimators of the proportions add up to one in the occupational sections, they are compositions that can be imprecise if the sample sizes are small. We fit a multivariate Fay–Herriot model to logratio transformations of the direct estimators of the proportions. Small area estimators of the proportions, entropies, and DIs are derived from the fitted model and the corresponding mean squared errors are estimated by parametric bootstrap. Several simulation experiments designed to analyze the behavior of the introduced model-based predictors are carried out. We give an application to Spanish Labour Force Survey data from 2022. The target is to investigate the state of sex occupational entropies and divergences in Spanish provinces.

Using isometric log-ratio in compositional data analysis for developing a groundwater pollution index

This study introduces a novel groundwater pollution index (GPI) formulated through compositional data analysis (CoDa) and robust principal component analysis (RPCA) to enhance groundwater quality assessment. Using groundwater quality monitoring data from sites impacted by the 2010–2011 foot-and-mouth disease outbreak in South Korea, CoDa uncovers critical hydrochemical differences between leachate-influenced and background groundwater. The GPI was developed by selecting key subcompositional parts (NH4+-N, Cl-, and NO3--N) using RPCA, performing the isometric log-ratio (ILR) transformation, and normalizing the results to environmental standards, thereby providing a more precise and accurate assessment of pollution. Validated against government criteria, the GPI has shown its potential as an alternative assessment tool, with its reliability confirmed by receiver operating characteristic curve analysis. This study highlights the essential role of CoDa, especially the ILR -transformation, in overcoming the limitations of traditional statistical methods that often neglect the relative nature of hydrochemical data. Our results emphasize the utility of the GPI in significantly advancing groundwater quality monitoring and management by addressing a methodological gap in the quantitative assessment of groundwater pollution.

Graph convolutional network for lithological classification and mapping using stream sediment geochemical data and geophysical data

Lithological mapping plays a vital role in acquiring a comprehensive understanding of subsurface rocks and geological structures. Traditional lithological mapping relies heavily on manual field surveys, with the accuracy of the mapping greatly constrained by the expertise of mappers, and thus this process is commonly slow and costly. In the present study, the graph convolutional network (GCN) method was proposed for efficient and automatic lithological mapping, using various geological survey data (e.g. geochemical and geophysical data). The GCN is not limited by the spatial distribution of sampling data, and hence it can be directly applied to the integrated resampling data that do not adhere to the regular spatial distribution patterns of Euclidean space. To evaluate the potential application of the GCN method in lithological classification and mapping, a case study was conducted in the Maqin region (China). Geochemical and geophysical data were utilized as the basis for lithological classification, as they can provide valuable information on geochemical compositions and physical properties of geological bodies, respectively. The study process was primarily divided into three parts. First, basic pre-processing was conducted on the geochemical data and the geophysical data, including centred log-ratio transformation and resampling. Subsequently, the resampled points were divided into training and testing sets in the proportion of 5.5:1. In the training set, 20% of the points are randomly selected as the validation set. Ultimately, using each resampled point as a node and the integrated geochemical and geophysical data as features, a GCN model and an undirected graph were constructed based on an adjacency distance of 1 km. The graph was employed by the constructed GCN model for lithological classification. The results reveal that the majority of lithological units can be accurately classified with an overall accuracy of 86%, indicating that the GCN model exhibits good applicability in lithological classification.

New soil weathering index based on compositional data analyses of silt to sand sized parent mineral assemblages of terra rossa soils

Weathering indices are widely used in soil science and some other environmental disciplines for a variety of purposes (e. g. for soil classification, in soil formation studies, or in wider palaeoclimatic research). Two main categories of weathering indices have been developed: geochemical and mineralogical. In both cases, an approach is developed that takes into account the compositional nature of the geochemical data. To date, however, the most commonly used weathering indices are geochemical indices that do not consider the geochemical data as compositional. Although the compositional approach minimises possible statistical errors, the weathering indices developed in both approaches, i.e., compositional and classical, axiomatically assume two things that are inaccurate in some cases. First, that soil composition is invariant with respect to the preferential translocation of clay-sized particles, i.e. lessivage, and second, that the selective sorting of minerals during transport does not affect the values of weathering indices. As a result, in some cases bulk analyses of mineralogical and geochemical content are an unreliable tool for determining weathering rates of soil. To overcome these difficulties, this study proposes a new weathering index (Wp-min) based on the orthonormal log ratio (olr) transformation of quantitative mineral data derived from silt- to sand-sized allochthonous mineral assemblages. The size fraction used ensures that only the parent mineral assemblages, which are not prone to translocation, are analysed. In this paper, olr transformation is done using the concept of balances. This enabled the construction of variables, which essentially are various mineral log ratios, with the desired properties, i.e. transport invariance and sensitivity to selective dissolution. In this way, undesirable effects, i.e. selective sorting and lessivage, which ultimately affect bulk analyses, are avoided, allowing a more accurate estimation of weathering in the soil profiles studied. The new index was validated by comparison with the geochemical W index, which was modified by consulting the standardized variation matrix prior to element selection. Ultimately, compared to the geochemical index, the new index was better able to characterise weathering in soils where intense lessivage was detected. The study was conducted on terra rossa soils, which have common source, mostly of aeolian origin.

Abstract 3927: Metabolic inhibition of BATF2 dampens type-I interferon-mediated immune sensing of cancer

Abstract Background: The type-I interferon (IFN-I) system appeared before adaptive immunity during evolution as the primary defense system for metazoans. Initiating cancers frequently disable this first line of defense before establishing T-cell exhaustion programs. However, the molecules linking IFN-I, danger signals, and unique metabolic conditions in the tumor microenvironment (TME) remain to be discovered. Methods: We performed centered log-ratio transformation on the cluster composition data using the R-package ALDEx2 to compare the clusters in single-cell data sets. IFN-I signatures were quantified using qPCR and ELISA. ChIP was performed to examine the H3K27me3 epigenetic marker on the promoters of BATF2 and IFN-I genes in bone marrow-derived macrophage (BMDM). For the cigarette smoke carcinogen-induced head and neck squamous cell carcinoma (HNSCC) model, experimental mice were fed with water containing 50 μg/mL of 4-NQO for 16 weeks and then switched to regular feeding water. We started monitoring 10 weeks post-treatment for the number, location, and diameter of the lesions. TME analysis was performed using single-cell RNA-Seq and multispectral imaging. Results: We analyzed three large independent cohorts of patients and found that BATF2 strongly correlated with the immunogenicity of HNSCC, with a stronger correlation coefficient than STING. We generated Batf2−/− BMDM and found that Batf2 deficiency rendered cells insensitive to STING agonists. However, the dsRNA-induced IFN-I activation was not affected by Batf2 deletion. BATF2 interacted with STING, promoting its phosphorylation and oligomerization. Glutamine is enriched in the TME. We found that glutamine inhibited the BATF2-STING pathway by increasing a repressive epigenetic marker, H3K27me3, on the promoters of BATF2 and IFNB1. The expression of BATF2 in cancer cells resulted in IFN-I- and γδ T-cell-dependent tumor rejection. Single-cell immune profiling showed that Batf2-expressing tumors contained significantly expanded Cxcl10+ myeloid cells and γδ T-cells. Utilizing the 4-NQO-induced HNSCC model, we found that BATF2 limited suppressive myeloid cells and expanded effector cells in the TME, serving as a potent tumor suppressor. Conclusion: BATF2 is a glutamine-responsive tumor suppressor that limits tumor initiation by engaging the STING-γδ T-cell pathway. Acknowledgments: This work was supported by NIH grants R01 DE026728, U01 DE029255, R01 DE031951, and U01 DE033330 Citation Format: Wang Gong, Hülya Taner, Yuesong Wu, Zaiye Li, Wanqing Cheng, Christopher Donnelly, Felipe Nör, Yumin He, Zackary Fitzsimonds, Jianqian Li, Haitao Wen, Steven Chinn, Yuying Xie, James Moon, Yu Leo Lei. Metabolic inhibition of BATF2 dampens type-I interferon-mediated immune sensing of cancer [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2024; Part 1 (Regular Abstracts); 2024 Apr 5-10; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2024;84(6_Suppl):Abstract nr 3927.

A New Discovery of Cu Mineralization in the North Qaidam, Tibet via Log-Ratio, Robust Factor Analysis, and Spectrum–Area Modeling

The Qinghai–Tibet Plateau, abundant in mineral resources, is a treasure trove for geological explorers. However, exploration has been hindered by the presence of dense vegetation, weathering layers, and desert cover, particularly in the North Qaidam region. As a result, there is an urgent need to develop efficient geochemical composition methods. In a study of stream sediment geochemical data from the Lüliangshan area of the North Qaidam, the log-ratio transformation was used for data processing, Robust Factor Analysis (RFA) was used for model construction, and the spectrum–area (S-A) model was used for anomaly separation. The outcomes identified two distinct groups of element combinations associated with mineralization. The first group consisted of Au + Ag + Pb + Sb + Hg linked to tectonic-altered rock-type Au (antimony) deposits, while the second group consisted of Cu + Zn + Co and was closely associated with Cu-Ni sulfide deposits. The S-A fractal filtering technique amplified weak anomalies and minimized the area of anomalies against strong backgrounds. The study successfully detected substantial Cu mineralization in the source areas of geochemical anomalies in the Lüliangshan region. Consequently, the log-ratio–RFA–S-A fractal model has been proven to be an effective combination of methods for identifying and extracting geochemical anomalies from stream sediment samples and for mineral exploration in covered areas.

Robust score matching for compositional data

The restricted polynomially-tilted pairwise interaction (RPPI) distribution gives a flexible model for compositional data. It is particularly well-suited to situations where some of the marginal distributions of the components of a composition are concentrated near zero, possibly with right skewness. This article develops a method of tractable robust estimation for the model by combining two ideas. The first idea is to use score matching estimation after an additive log-ratio transformation. The resulting estimator is automatically insensitive to zeros in the data compositions. The second idea is to incorporate suitable weights in the estimating equations. The resulting estimator is additionally resistant to outliers. These properties are confirmed in simulation studies where we further also demonstrate that our new outlier-robust estimator is efficient in high concentration settings, even in the case when there is no model contamination. An example is given using microbiome data. A user-friendly R package accompanies the article.

Digital Mapping of Soil Particle Size Fractions in the Loess Plateau, China, Using Environmental Variables and Multivariate Random Forest

Soil particle size fractions (PSFs) are important properties for understanding the physical and chemical processes in soil systems. Knowledge about the distribution of soil PSFs is critical for sustainable soil management. Although log-ratio transformations have been widely applied to soil PSFs prediction, the statistical distribution of original data and the transformed data given by log-ratio transformations is different, resulting in biased estimates of soil PSFs. Therefore, multivariate random forest (MRF) was utilized for the simultaneous prediction of soil PSFs, as it is able to capture dependencies and internal relations among the three components. Specifically, 243 soil samples collected across the Loess Plateau were used. Meanwhile, Landsat data, terrain attributes, and climatic variables were employed as environmental variables for spatial prediction of soil PSFs. The results depicted that MRF gave satisfactory soil PSF prediction performance, where the R2 values were 0.62, 0.53, and 0.73 for sand, silt, and clay, respectively. Among the environmental variables, nighttime land surface temperature (LST_N) presented the highest importance in predicting soil PSFs in the Loess Plateau, China. Maps of soil PSFs and texture were generated at a 30 m resolution, which can be utilized as alternative data for soil erosion management and ecosystem conservation.

A novel point process model for neuronal spike trains

Point process provides a mathematical framework for characterizing neuronal spiking activities. Classical point process methods often focus on the conditional intensity function, which describes the likelihood at any time point given its spiking history. However, these models do not describe the central tendency or importance of the spike train observations. Based on the recent development on the notion of center-outward rank for point process, we propose a new modeling framework on spike train data. The new likelihood of a spike train is a product of the marginal probability on the number of spikes and the probability of spike timings conditioned on the same number. In particular, the conditioned distribution is calculated by adopting the well-known Isometric Log-Ratio transformation. We systematically compare the new likelihood with the state-of-the-art point process likelihoods in terms of ranking, outlier detection, and classification using simulations and real spike train data. This new framework can effectively identify templates as well as outliers in spike train data. It also provides a reasonable model, and the parameters can be efficiently estimated with conventional maximum likelihood methods. It is found that the proposed likelihood provides an appropriate ranking on the spike train observations, effectively detects outliers, and accurately conducts classification tasks in the given data.