Binary Matrix Research Articles

Reliable Crops Classification Using Limited Number of Sentinel-2 and Sentinel-1 Images

The study presents the analysis of the possible use of limited number of the Sentinel-2 and Sentinel-1 to check if crop declarations that the EU farmers submit to receive subsidies are true. The declarations used in the research were randomly divided into two independent sets (training and test). Based on the training set, supervised classification of both single images and their combinations was performed using random forest algorithm in SNAP (ESA) and our own Python scripts. A comparative accuracy analysis was performed on the basis of two forms of confusion matrix (full confusion matrix commonly used in remote sensing and binary confusion matrix used in machine learning) and various accuracy metrics (overall accuracy, accuracy, specificity, sensitivity, etc.). The highest overall accuracy (81%) was obtained in the simultaneous classification of multitemporal images (three Sentinel-2 and one Sentinel-1). An unexpectedly high accuracy (79%) was achieved in the classification of one Sentinel-2 image at the end of May 2018. Noteworthy is the fact that the accuracy of the random forest method trained on the entire training set is equal 80% while using the sampling method ca. 50%. Based on the analysis of various accuracy metrics, it can be concluded that the metrics used in machine learning, for example: specificity and accuracy, are always higher then the overall accuracy. These metrics should be used with caution, because unlike the overall accuracy, to calculate these metrics, not only true positives but also false positives are used as positive results, giving the impression of higher accuracy. Correct calculation of overall accuracy values is essential for comparative analyzes. Reporting the mean accuracy value for the classes as overall accuracy gives a false impression of high accuracy. In our case, the difference was 10–16% for the validation data, and 25–45% for the test data.

Venation pattern and micromorphology of schizocarps of Gaya and related genera within Malveae (Malvaceae)

Few studies have reported the venation pattern and micromorphology of Malveae schizocarpic fruits. These reproductive structures are present in Gaya and other genera of the tribe and may prove useful for solving taxonomic problems and understanding phylogenetic relationships within. We analyzed the mericarp and the seeds of 13 Brazilian species of Gaya and 11 species of related genera to provide useful comparative features for their taxonomy and phylogeny. The samples were subjected to the herborization reversal procedure and followed methods of scanning electron microscopy, light microscopy and diaphanization. The studied species of Gaya were compared on a binary matrix of characters analyzed through UPGMA. Within Gaya, the mericarp constriction corresponds to the dehiscence lines, which may occur at the basal portion or ventrally. Two to three veins in contact with the columella (central structure to which the mericarps are attached) were observed in this region. At the upper ventral margin, we observed only one vein connecting with the columella. The exocarp exhibited simple, biradiate, stellate and multiradiated trichomes. Capitate glandular trichomes were also observed. Species from related genera (Briquetia, Callianthe, Herissantia and Hochreutinera) showed stellate trichomes. Glandular and tector trichomes were observed in Gaya seeds, while other genera exhibited only simple trichomes. Venation pattern and micromorphological characters of mericarps, as well as novel seed trichome types here reported support of Gaya support the delimitation of genera studied and highlight the isolated position of Gaya within Malvaeae.

Synthesis of a Sparse 2D-Scanning Array using Particle Swarm Optimization for Side-Lobe Reduction

Synthesis of sparse arrays is a promising area of research for a wide range of applications including radar and millimeter-wave wireless communication. The design goal of array thinning problems is to reduce the number of elements of an array without significantly affecting its performance. This work presents a technique for synthesizing a sparse phased-array antenna from a 16×16 uniform rectangular array (URA). The proposed approach reduces the number of elements by 50% without any significant increase in the peak sidelobe level (PSLL) for all possible scan angles in the azimuthal and elevation plans within a finite range of scan angles. The synthesis includes an artificial neural network (ANN) model for estimation of the excitation weights of the URA for a given scan-angle. The weights of the sparse array are computed by the Hadamard product of the weight matrix of the URA with a binary matrix that is obtained using particle swarm optimization (PSO) to minimize the PSLL.

A 2D model for prediction of nanoparticle distribution and microstructure evolution during solidification of metal matrix nanocomposites

A 2D numerical model to predict the microstructure evolution in binary alloy matrix nanocomposites is presented. The main novelty of the model is that it captures the effect of nanoparticles on binary alloy dendrite growth and microstructure evolution. Diffusion based transport of nanoparticles is solved using a nanoparticle concentration defined in a similar way to an alloying element. Engulfment or rejection of nanoparticles at the solidification interface is implemented by defining a partition coefficient. The diffusivity of nanoparticles is calculated based on nanoparticle movement due to Brownian motion. The governing equations for binary alloy solidification are formulated using the volume averaged enthalpy method with a probabilistic nucleation model for grain nucleation. The effect of nanoparticles on solidification is incorporated through modifications of the solute diffusion field and interfacial energy. The developed model is used for predicting the effect of nanoparticle concentration and nanoparticle radius on dendrite growth and microstructure evolution. From single dendrite study, it is observed that the dendrite size decreases with increase in nanoparticle concentration while it increases with increase in nanoparticle radius. During microstructure evolution, increased non-homogeneity in nanoparticle distribution is observed with increase in nanoparticle concentration and more uniform distribution is observed with increase in nanoparticle radius.

Deterministic Constructions of Compressed Sensing Matrices From Unitary Geometry

Compressed sensing is an emerging theory of signal processing and it has wide applications in many frontier fields. The construction of the measurement matrices is still a central problem in compressed sensing. In this paper, two types of deterministic constructions of binary measurement matrices are presented via unitary geometry. Then, the lower bounds of the spark of unitary geometry measurement matrices are theoretically analyzed, and an asymptotic comparison between unitary geometry measurement matrices and projective geometry measurement matrices is given via the worst-case recovery capability. After that, a clipping-embedding operation is proposed for binary matrices to generate measurement matrices with more sizes, which can strongly extend the applicability of the deterministic binary matrices in practice. Finally, simulation results demonstrate that the performance of our measurement matrices is comparable to, sometimes even better than, that of the corresponding Gaussian random matrices under OMP and BP.

Sum-Rank BCH Codes and Cyclic-Skew-Cyclic Codes

In this work, cyclic-skew-cyclic codes and sum-rank BCH codes are introduced. Cyclic-skew-cyclic codes are characterized as left ideals of a suitable non-commutative finite ring, constructed using skew polynomials on top of polynomials (or vice versa). Single generators of such left ideals are found, and they are used to construct generator matrices of the corresponding codes. The notion of defining set is introduced, using pairs of roots of skew polynomials on top of poynomials. A lower bound (called sum-rank BCH bound) on the minimum sum-rank distance is given for cyclic-skew-cyclic codes whose defining set contains certain consecutive pairs. Sum-rank BCH codes, with prescribed minimum sum-rank distance, are then defined as the largest cyclic-skew-cyclic codes whose defining set contains such consecutive pairs. The defining set of a sum-rank BCH code is described, and a lower bound on its dimension is obtained. Thanks to it, tables are provided showing that sum-rank BCH codes beat previously known codes for the sum-rank metric for binary 2 ×2 matrices (i.e., codes whose codewords are lists of 2 ×2 binary matrices, for a wide range of list lengths that correspond to the code length). Finally, a decoder for sum-rank BCH codes up to half their prescribed distance is obtained.

Internal Colonialism and the Discipline of Comparative Literature in Iran

Abstract This essay studies the history, current, and future status of the discipline of comparative literature in Iran. It compares the theoretical norms of contemporary comparative literature to the Pre-modern Perso-Islamic notion of “comparison,” which has been theorized in Iran and the Arab World as the Arabic, Islamic, and Iranian schools of comparative literature. The article highlights profound institutional and canonical Perso-Shi’a centrism in Iranian academia, and shows how the discipline of comparative literature has been used as a vehicle for transnationalism of this Perso-Shi’a centrism that has manifested in “Persianate World” in the context of European and North American academia. Marshall Hodgson’s 1960s neologism “Persianate World” has been placed with the paradigm shifts ushered in by the linguistic and cultural turns of the 1970s, the postcolonial scholarship that grew from Edward Said’s Orientalism in the late 1990s, and Sheldon Pollock’s formulation of a ‘Sanskrit cosmopolis’ in the 21st century. The article explains how the Persianate comparatists, under the banner of postcolonial studies, not only erased the experience of the subaltern and internally colonialized non-Persians of Iran in favor of the Middle Eastern states in a binary matrix (Western Imperialism versus a “colonialized” Islamic world), but also represents an unrealistic and exaggerated picture of the discipline to Western readers. The article further maps the conversations within the postcolonial Middle East about “internal colonialism,” as an analytic tool for thinking about operations and interlocking systems of power in the Middle East and abroad, here applied to the discipline of comparative literature for the first time.

Translation of Six Tuple Grade-1 Braille Alphabet to English Alphabet

Braille is the language used by visually impaired persons. Braille language comprises of collection of Braille cells which are embossed on a metal plate. Maintaining these bulky metal plates and distributing them to other parts of the world is a challenging task. This paper proposes a new technique of translating the Braille Cells embossed on plate to a natural language English character which can be easily distributed over network to make it globally accessible. Initially Braille documents are scanned, and preprocessing techniques like adaptive histogram and Laplacian filters are applied to augment the dots by eliminating the noise. The existence of dot pattern in every cell is detected with a Threshold and transformed to sequence of Binary matrix. A cell information is translated to 3x2 matrix with binary values of 0’s and 1’s representing absence and presence of dots in a cell. Convolutional Neural Network is used for feature extraction and Classification and Regression Trees (CART) classifier is utilized for recognize the character.

Assessment of the Thermal Effect on the Surface of Metal Structural Materials on the Stability of Laser-Induced Codes Readability

This work represents the characterization of materials surface before and after laser processing with macrophotography, optical metallography, and scanning electron microscopy before and after thermal exposure. The factors influencing the reliability of the laser-induced code readability have been determined as color and contrast. The range of stability of the code readability under thermal influence on the structural materials under study was determined, which allows improving the reliability of the laser-induced marking codes readability. The research objects in this paper were samples of the following materials: alloys based on copper, aluminum, and iron with laser-induced codes of various types applied on the surface. This work aimed to research the stability of laser-induced codes readability after thermal exposure using macrophotography, optical metallography and scanning electron microscopy on structural materials of various purposes before and after laser processing (when forming a binary matrix code). The research results obtained and presented in this article on the stability of laser-induced codes reading under thermal action on structural materials can be used in different fields of industry, when marking products of heavy, general, medium, and precision engineering, as well as for marking metal products and blank parts. The results of this research are also planned to be used for further analysis of the occurring damage, leading to reading errors due to mechanical and chemical influences. It is planned to evaluate the limit values of the parameters that determine the degree of degradation at which the encoding will be considered to have lost the recognizing ability. Requirements for the quality and permissible code damage will also be developed to ensure their reliable identification.

Studying the History of Tumor Evolution from Single-Cell Sequencing Data by Exploring the Space of Binary Matrices.

Single-cell sequencing (SCS) data have great potential in reconstructing the evolutionary history of tumors. Rapid advances in SCS technology in the past decade were followed by the design of various computational methods for inferring trees of tumor evolution. Some of the earliest methods were based on the direct search in the space of trees with the goal of finding the maximum likelihood tree. However, it can be shown that instead of searching directly in the tree space, we can perform a search in the space of binary matrices and obtain maximum likelihood tree directly from the maximum likelihood matrix. The potential of the latter tree search strategy has recently been recognized by different research groups and several related methods were published in the past 2 years. Here we provide a review of the theoretical background of these methods and a detailed discussion, which are largely missing in the available publications, of the correlation between the two tree search strategies. We also discuss each of the existing methods based on the search in the space of binary matrices and summarize the best-known single-cell DNA sequencing data sets, which can be used in the future for assessing performance on real data of newly developed methods.

Identification of Somatic Mutation-Driven Immune Cells by Integrating Genomic and Transcriptome Data.

Tumor somatic mutations in protein-coding regions may generate neoantigens which may trigger antitumor immune cell response. Increasing evidence supports that immune cell response may profoundly influence tumor progression. However, there are no calculated tools to systematically identify immune cells driven by specific somatic mutations. It is urgent to develop a calculated method to comprehensively detect tumor-infiltrating immune cells driven by the specific somatic mutations in cancer. We developed a novel software package (SMDIC) that enables the automated identification of somatic mutation-driven immune cell. SMDIC provides a novel pipeline to discover mutation-specific immune cells by integrating genomic and transcriptome data. The operation modes include inference of the relative abundance matrix of tumor-infiltrating immune cells, detection of differential abundance immune cells with respect to the gene mutation status, conversion of the abundance matrix of significantly dysregulated cells into two binary matrices (one for upregulated and one for downregulated cells), identification of somatic mutation-driven immune cells by comparing the gene mutation status with each immune cell in the binary matrices across all samples, and visualization of immune cell abundance of samples in different mutation status for each gene. SMDIC provides a user-friendly tool to identify somatic mutation-specific immune cell response. SMDIC may contribute to understand the mechanisms underlying anticancer immune response and find targets for cancer immunotherapy. The SMDIC was implemented as an R-based tool which was freely available from the CRAN website https://CRAN.R-project.org/package=SMDIC.

Topology-based sparsification of graph annotations.

MotivationSince the amount of published biological sequencing data is growing exponentially, efficient methods for storing and indexing this data are more needed than ever to truly benefit from this invaluable resource for biomedical research. Labeled de Bruijn graphs are a frequently-used approach for representing large sets of sequencing data. While significant progress has been made to succinctly represent the graph itself, efficient methods for storing labels on such graphs are still rapidly evolving.ResultsIn this article, we present RowDiff, a new technique for compacting graph labels by leveraging expected similarities in annotations of vertices adjacent in the graph. RowDiff can be constructed in linear time relative to the number of vertices and labels in the graph, and in space proportional to the graph size. In addition, construction can be efficiently parallelized and distributed, making the technique applicable to graphs with trillions of nodes. RowDiff can be viewed as an intermediary sparsification step of the original annotation matrix and can thus naturally be combined with existing generic schemes for compressed binary matrices. Experiments on 10 000 RNA-seq datasets show that RowDiff combined with multi-BRWT results in a 30% reduction in annotation footprint over Mantis-MST, the previously known most compact annotation representation. Experiments on the sparser Fungi subset of the RefSeq collection show that applying RowDiff sparsification reduces the size of individual annotation columns stored as compressed bit vectors by an average factor of 42. When combining RowDiff with a multi-BRWT representation, the resulting annotation is 26 times smaller than Mantis-MST.Availability and implementationRowDiff is implemented in C++ within the MetaGraph framework. The source code and the data used in the experiments are publicly available at https://github.com/ratschlab/row_diff.

Genome Instability-Derived Genes Are Novel Prognostic Biomarkers for Triple-Negative Breast Cancer.

BackgroundTriple-negative breast cancer (TNBC) is an aggressive disease. Recent studies have identified genome instability-derived genes for patient outcomes. However, most of the studies mainly focused on only one or a few genome instability-related genes. Prognostic potential and clinical significance of genome instability-associated genes in TNBC have not been well explored.MethodsIn this study, we developed a computational approach to identify TNBC prognostic signature. It consisted of (1) using somatic mutations and copy number variations (CNVs) in TNBC to build a binary matrix and identifying the top and bottom 25% mutated samples, (2) comparing the gene expression between the top and bottom 25% samples to identify genome instability-related genes, and (3) performing univariate Cox proportional hazards regression analysis to identify survival-associated gene signature, and Kaplan–Meier, log-rank test, and multivariate Cox regression analyses to obtain overall survival (OS) information for TNBC outcome prediction.ResultsFrom the identified 111 genome instability-related genes, we extracted a genome instability-derived gene signature (GIGenSig) of 11 genes. Through survival analysis, we were able to classify TNBC cases into high- and low-risk groups by the signature in the training dataset (log-rank test p = 2.66e−04), validated its prognostic performance in the testing (log-rank test p = 2.45e−02) and Molecular Taxonomy of Breast Cancer International Consortium (METABRIC) (log-rank test p = 2.57e−05) datasets, and further validated the predictive power of the signature in five independent datasets.ConclusionThe identified novel signature provides a better understanding of genome instability in TNBC and can be applied as prognostic markers for clinical TNBC management.

Multi-image Reorganization Encryption Based on S-L-F Cascade Chaos and Bit Scrambling

Aiming at the problems of small value range of a single chaotic parameter, low sequence chaos, and transient effects, a composite chaotic system of cascaded Sine-Sine mapping, Logistic chaos and generalized third-order Fibonacci is proposed (S-L-F). The new system is highly sensitive to initial values, the maximum spectral entropy of the generated sequence can reach 0.95, and the value range of the parameter x is expanded to [0,4] compared with the traditional Logistic, indicating that the new system is suitable for generating pseudo-random sequences for image encryption. For the problem that the traditional multi-image encryption scheme can only encrypt images of the same type and size, the practicability is poor, and a multi-image encryption scheme based on image reorganization and biting is proposed. The algorithm recombines any number, different sizes and different types of images into a three-dimensional matrix, converts it into a binary matrix, performs bit-level scrambling and surface cyclic scrambling, and then restores the scrambling matrix to decimal, and the chaotic sequence performs exclusive-or diffusion, and completes simultaneous encryption at one time, which greatly improves the encryption efficiency and scope of application. The NPCR of the ciphertext image is 0.9961, and the UACI is 0.3345, which proves that the ciphertext image can effectively resist the difference attack. The information entropy is greater than 7.999, which can effectively resist attacks. It has certain application value in image information security. Experimental analysis shows that the algorithm has high security and strong practicability.

Analysis of trigonometric chaotic sequence by proposing an index-based bit level scrambling image encryption

Chaotic map is applied to numerous research fields, such as encryption of data and information. In this paper, a novel bit-level scrambling encryption based on three-dimensional trigonometric chaotic sequence is proposed. This encryption is operated on grayscale images because pixel value of which is between 0 and 255, which can be converted into 8-bit binary number, then each image can be converted into eight binary images. Combine binary images a three-dimensional matrix as input, which have three index values width, height, and page. Generate three numbers by the chaotic sequence to locate the value of the matrix need to be changed. After the value is located, change its value 0 to 1 or 1 to 0, this is the basic encryption scheme. This paper makes a three-dimensional binary matrix with 32 images to test the encryption scheme, and get average value of images’ number of pixel change rate at 0.9603, unified averaged change intensity at 31.27%, information entropy at 7.9891, also the histograms. And correlation coefficients of each pixel from three directions are small. There are comparisons with other encryptions too. The key space of this encryption is more than 10[Formula: see text].

Compact and High-Speed Hsiao-Based SEC-DED Codec for Cache Memory

Recently, there have been continuous rising interests of multi-bit error correction codes (ECCs) for protecting memory cells from soft errors which may also enhance the reliability of memory systems. The single error correction and double error detection (SEC-DED) codes are generally employed in many high-speed memory systems. In this paper, Hsiao-based SEC-DED codes are optimized based on two proposed optimization algorithms employed in parity check matrix and error correction logic. Theoretical area complexity of SEC-DED codecs require maximum 49.29%, 18.64% and 49.21% lesser compared to the Hsiao codes [M. Y. Hsiao, A class of optimal minimum odd-weight-column SEC-DED codes, IBM J. Res. Dev. 14 (1970) 395–401], Reviriego et al. codes [P. Reviriego, S. Pontarelli, J. A. Maestro and M. Ottavi, A method to construct low delay single error correction codes for protecting data bits only, IEEE Trans. Comput.-Aided Des. Integr. Circuits Syst. 32 (2013) 479–483] and Liu et al. codes [S. Liu, P. Reviriego, L. Xiao and J. A. Maestro, A method to recover critical bits under a double error in SEC-DED protected memories, Microelectron. Reliab. 73 (2017) 92–96], respectively. Proposed codec is designed and implemented both in field programmable gate array (FPGA) and ASIC platforms. The synthesized SEC-DED codecs need 31.14% lesser LUTs than the original Hsiao code. Optimized codec is faster than the existing related codec without affecting its power consumption. These compact and faster SEC-DED codecs are employed in cache memory to enhance the reliability.

Abstract 2240: Genomic profiling of smoldering multiple myeloma classifies distinct molecular groups

Abstract Background: Multiple Myeloma (MM) is an incurable plasma cell malignancy with significant genomic heterogeneity. It is usually preceded by the asymptomatic stage known as smoldering multiple myeloma (SMM). SMM patients have a 10% annual risk of progression to MM. Genomic alterations that are observed in SMM patients include chromosomal gains and losses, translocations, and point mutations. However, current SMM risk models rely solely on clinical markers that do not accurately capture the progression risk. While incorporating some genomic biomarkers improves prediction, using all MM genomic features to comprehensively stratify patients may increase the precision of risk models. Methods: We obtained a total of 214 patients' samples at SMM diagnosis in the US and Europe. We performed whole exome sequencing on 166 tumors; of these, RNA sequencing was performed on 100. Targeted capture with a MM gene panel was done on an additional 48 tumors. We identified subgroups using binarized DNA features and performing consensus binary non-negative matrix factorization. Results: We identified six clusters (C1-C6) with the following features: four with a hyperdiploidy (HD) (>48 chromosomes) and two with IgH translocations. These subgroups have unique transcriptomic profiles overlapping with known MM signatures and biological pathways. One of the clusters harboring translocation (11;14), which we call C4-CCND1, was enriched with the previously defined CD-2 MM signature that uniquely expresses B cell markers CD20 and CD79A; shows upregulation of CCND1 and E2F7; and is enriched with pathways like DNA replication, heme metabolism, and NFkB signaling. The C3-MS_MF cluster with the IgH translocations (4;14) and (14;16) shows downregulation of ribosomal genes, TRAF2, and DUSP2. The MYC oncogene was highly expressed in the four HD clusters: C1-HD_NRAS, C2-HD_MAFB, C5-HD_KRAS, and C6-HD_1q (BH-P = 0.037). The clusters also showed different outcomes in terms of time to progression (TTP) to active MM (P = 0.005). Median TTP for patients in C2-HD_MAFB, C3-MS_MF, and C5-HD_KRAS was 3.7, 2.6, and 2.2 years, respectively; TTP for C1-HD_NRAS, C4-CCND1, and C6-HD_1q was 4.3, 11, and not reached, respectively. In multivariate analysis, C2-HD_MAFB, C3-MS_MF, and C5-HD_KRAS were independent predictors of progression after accounting for the clinical risk stage. Moreover, the odds of having evolving hemoglobin and monoclonal protein levels in these three clusters were 3.5 and 12.3 times higher than the other clusters, respectively (P = 0.01 and 0.002). Conclusion: We identified six distinct SMM molecular groups with corresponding transcription profiles and dysregulated pathways. These groups have different progression risks to active MM, with three groups being independent predictors of progression. Our results underscore the importance of molecular classification in MM to better understand and target various tumor vulnerabilities. Citation Format: Shankara Anand, Mark Bustoros, François Aguet, Romanos Sklavenitis-Pistofidis, Robert Redd, Benny Zhitomirsky, Andrew J. Dunford, Yu-Tzu Tai, Selina J. Chavda, Cody Boehner, Carl J. Neuse, Tineke Casneuf, Lorenzo Trippa, Chip Stewart, Kwee Yong, Irene Ghobrial, Gad Getz. Genomic profiling of smoldering multiple myeloma classifies distinct molecular groups [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2021; 2021 Apr 10-15 and May 17-21. Philadelphia (PA): AACR; Cancer Res 2021;81(13_Suppl):Abstract nr 2240.

Abstract LB258: Branching cancer evolution biases the detection of epistatic interactions between driver mutations

Abstract Cancer genomics datasets can be leveraged to decipher evolutionary dependencies, an important frontier for precision oncology. Phylogenetic analyses time mutations as occurring early (clonal) or late (subclonal) in the life history of a tumour. A separate class of computational tools aims to systematically detect mutually exclusive (ME) relationships, revealing evolutionary dependencies including synthetic lethal or functionally redundant interactions between mutations. Existing tools for ME detection take binary patient-level mutation matrices as input, ignoring the branching clonal architecture that characterizes most cancer genomes. This may result in a bias to detect clonal mutations. Moreover, ME interactions between subclonal mutations in separate lineages of a tumor's phylogenetic tree may be overlooked. Here, we examine the influence of mutation clonality on ME detection in a pan-cancer analysis. Whole exome-sequencing data was accessed from the TCGA data portal for 3,545 patients from nine major cancer-types. Nonsense and missense mutations in driver genes were filtered. PyClone was used to estimate the cancer cell fraction (CCF) for each variant. To detect significant (P<0.05) ME interactions, the DISCOVER and TiMEx tools were used.First, focusing on patients with lung adenocarcinoma (LUAD), the median CCF value for each gene across the cohort was used as a summary clonality metric, indicating that 58% (106/182 genes) of the driver genes assessed in LUAD tend to be clonally mutated (median CCF = 1). Across nine cancer-types, on average 58% (range=25-68%) of mutations were classified as clonal. In parallel, the DISCOVER and TiMEx tools for ME detection were applied to the LUAD dataset. Clonally mutated genes comprised the majority of the significant interactions detected in LUAD tumours for both tools: DISCOVER=89% (107 of 120), TiMEx=77% (127 of 166). This bias towards detecting clonal mutations was the general pattern across cancer-types: DISCOVER median=81% (range = 38-93%); TiMEx median=77% (range=48-81%). Using the DISCOVER tool in the LUAD cohort, the tendency of a gene to be clonally mutated exhibited a highly significant positive correlation with the proportion of significant findings in ME analysis (P=0.00685). The TiMEx tool, which uses a different underlying model and set of assumptions, also identified significant positive associations in LUAD (P=0.0251), demonstrating that the influence of mutation clonality on ME detection is not specific to a particular tool. Overall, significant positive correlations between mutation clonality and the number of statistically significant ME interactions were found for 78% (7/9) of cancer-types using DISCOVER, and 44% (4/9) using TiMEx. Our results suggest mutation clonality might represent an unaddressed confounding factor for the detection of significant ME interactions across cancer-types. Addressing this bias, in future, could help define a core set of evolutionary dependencies. Citation Format: Dhruva Biswas, Nicolai J. Birkbak, Javier Herrero, Charles Swanton. Branching cancer evolution biases the detection of epistatic interactions between driver mutations [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2021; 2021 Apr 10-15 and May 17-21. Philadelphia (PA): AACR; Cancer Res 2021;81(13_Suppl):Abstract nr LB258.

Abstract 2279: Improving cell type and state classification in CITE-seq using probabilistic models and hierarchical cell ID matrix

Abstract Accurate classification of cells in a population is important for most immune phenotyping applications. Surface protein marker expression measured with multiplexed platforms like Flow Cytometry and sequential gating strategies are typically employed to classify cells. However, new single-cell RNA sequencing (scRNA-seq) methods that measure RNA expression alone or Cellular Indexing of Transcriptomes and Epitopes by sequencing (CITE-seq) methods that concurrently measure both surface protein and RNA expression in single cells are providing new data and opportunities to develop methods both to classify cells and to assess their states of activation. Traditional manual gating methods using biaxial plots, where one designates a threshold separating a bimodal distribution through visual approximation, are subjective and can vary widely between individuals. Instead of manual, sequential gating for cell classification, we have implemented a probabilistic Gaussian Mixture Model (GMM) and K-means clustering to separate cell populations based on a library of marker thresholds for a variety of cell types present in human peripheral blood mononuclear cells (PBMCs) using a CITE-seq ADT expression matrix that has been preprocessed with centered-log-ratio (CLR) normalization. Furthermore, we coupled this GMM method with a hierarchical approach for cell typing. We have developed a Cell ID Matrix that outlines over 40 different immune cell types using consensus markers applicable to immune cell types across various studies. In this typing method, major cell types are identified following positive and negative marker expression as described in the Cell ID Matrix. Hierarchical gating of cells can then be used to identify subpopulations in a principled manner. The information in the Cell ID matrix is input as a binary matrix for each hierarchy level and is matched to the binarized expression values for classifying the cells. This approach can also be extended to CITE-seq datasets from disassociated tumor cells (DTCs). We conclude that automated cell gating with CITE-seq data may enhance the classification of numerous cell types in addition to significantly decreasing the amount of human intervention and time allocated for gating cells. Identifying cell states is a work in progress and will supplement this technique in the future. Citation Format: Lila Fakharzadeh, Julien Tessier, Shannon McGrath, Emma Wang, Angelique Biancotto, Alexei Protopopov, Jack Pollard, Joon Sang Lee. Improving cell type and state classification in CITE-seq using probabilistic models and hierarchical cell ID matrix [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2021; 2021 Apr 10-15 and May 17-21. Philadelphia (PA): AACR; Cancer Res 2021;81(13_Suppl):Abstract nr 2279.

Analysis of multi-class classification performance metrics for remote sensing imagery imbalanced datasets

Remote sensing imaging datasets for classification generally present high levels of imbalance between classes of interest. This work presented a study of a set of performance evaluation metrics for an imbalance dataset. In this work, a support vector machine (SVM) was used to perform the classification of seven classes of interest in a popular dataset called Salinas-A. The performance evaluation of the classifier was performed using two types of metrics: 1) Metrics for multi-class classification, and 2) Metrics based on the binary confusion matrix. In the results, a comparison of the scores of each metric is developed, some being more optimistic than others due to the bias that they present given the imbalance. In addition, our case study helps to conclude that the Matthews correlation coefficient (MCC) presents the lowest bias in imbalanced cases and is regarded to be robust metric. These results can be extended to any imbalanced dataset taking into account the equations developed by Luque.