Biology subjects

Kotlarz, K.

Publications and source records attributed to Kotlarz, K..

6 recordsLinked to original sources

bioRxiv2024

Unravelling traits with complex inheritance mode: a gene panel perspective

Although gene panels for complex phenotypes target genes contributing significantly to phenotypic variation, identifying causal mutations remains challenging. We aimed to define the process of analysing gene panel data to uncover genomic insights for traits with complex inheritance. The implemented procedure included the exploration of (i) the association between phenotypic variability and the variability of SNPs within genes, (ii) the association attributed to specific variants, (iii) causality potential by analysing the genomic context in the vicinity of the significantly associated SNPs (iii_a) either through clustering - to identify variants with genotypic patterns similar to the significant SNPs, or (iii_b) through smoothing the set of original P values to incorporate the information on the impact of the particular mutations, as expressed by the SNP-specific PhRED-CADD scores. This procedure was applied to the data from the obesity gene panel. No significant gene-level associations were found. However, the SNP-level results were promising in narrowing down the initially available set of variants to determine the true causal mutations. We introduced a procedure for analysing gene panel data to identify potentially causal mutations for phenotypes with complex inheritance patterns. The procedure allows for narrowing down the originally available set of variants offered by a gene panel, consisting of genes important for the phenotype under study, to SNPs that have a statistically and genomically validated potential to be causal in modifying protein structure.

bioinformatics↗

bioRxiv2024

Approaches to dimensionality reduction for ultra-high dimensional models

The rapid advancement of high-throughput sequencing technologies has revolutionised genomic research by providing access to large amounts of genomic data. However, the most important disadvantage of using Whole Genome Sequencing (WGS) data is its statistical nature, the so-called p>>n problem. This study aimed to compare three approaches of feature selection allowing for circumventing the p>>n problem, among which one is a novel modification of Supervised Rank Aggregation (SRA). The use of the three methods was demonstrated in the classification of 1,825 individuals representing the 1000 Bull Genomes Project to 5 breeds, based on 11,915,233 SNP genotypes from WGS. In the first step, we applied three feature (i.e. SNP) selection methods: the mechanistic approach (SNP tagging) and two approaches considering biological and statistical contexts by fitting a multiclass logistic regression model followed by either 1-dimensional clustering (1D-SRA) or multi-dimensional feature clustering (MD-SRA) that was originally proposed in this study. Next, we perform the classification based on a Deep Learning architecture composed of Convolutional Neural Networks. The classification quality of the test data set was expressed by macro F1-Score. The SNPs selected by SNP tagging yielded the least satisfactory results (86.87%). Still, this approach offered rapid computing times by focussing only on pairwise LD between SNPs and disregarding the effects of SNP on classification. 1D-SRA was less suitable for ultra-high-dimensional applications due to computational, memory and storage limitations, however, the SNP set selected by this approach provided the best classification quality (96.81%). MD-SRA provided a very good balance between classification quality (95.12%) and computational efficiency (17x lower analysis time and 14x lower data storage), outperforming other methods. Moreover, unlike SNP tagging, both SRA-based approaches are universal and not limited to feature selection for genomic data. Our work addresses the urgent need for computational techniques that are both effective and efficient in the analysis and interpretation of large-scale genomic datasets. We offer a model suitable for the classification of ultra-high-dimensional data that implements fusing feature selection and deep learning techniques.

bioinformatics↗

bioRxiv2024

Exploring the impact of sequence context on errors in SNP genotype calling with Whole Genome Sequencing data using AI-based autoencoder approach

A critical step in the analysis of WGS data is variant calling. Despite its importance, variant calling is prone to errors. Our study investigated the association between incorrect SNP and variant quality metrics and nucleotide context. In our study, incorrect SNPs were defined in twenty Holstein-Friesian cows by comparing their SNPs genotypes identified by whole genome sequencing with the IlluminaNovaSeq6000 and the EuroGMD50K genotyping microarray. The data set was divided into the correct set of SNPs (666,333 SNPs) and the incorrect set of SNPs (4,557 SNPs). The training data set consisted of only the correct SNPs, while the test data set contained a balanced mix of all the incorrectly and correctly called SNPs. An autoencoder was constructed to identify systematically incorrect SNPs that were marked as outliers by a one-class support vector machine and isolation forest algorithms. The results showed that 59.53% ({+/-}0.39%) of the incorrect SNPs had systematic patterns, with the remainder being random errors. The frequent occurrence of the CGC trimer was due to mislabeling a call for C. Incorrect T instead A call was associated with the presence of T in the neighboring downstream position. These errors may arise due to the fluorescence patterns of nucleotide labelling.

bioinformatics↗

bioRxiv2024

Nextflow vs. plain Bash: Different Approaches to the Parallelisation of SNP Calling from the Whole Genome Sequence Data

This study compared computational approaches to parallelisation of an SNP calling workflow. Data comprised DNA from five Holstein-Friesian cows sequenced with the Illumina platform. The pipeline consisted of quality control, alignment to the reference genome, post-alignment, and SNP calling. Three approaches to parallelisation were compared: (i) a plain Bash script in which a pipeline for each cow was executed as separate processes invoked at the same time, (ii) a Bash script wrapped in a single Nextflow process, and (iii) a Nextflow script with each component of the pipeline defined as a separate process. The results demonstrated that on average, the multi-process Nextflow script performed 15% to 27% faster depending on the number of assigned threads, with the biggest execution time advantage over the plain Bash approach observed with 10 threads. In terms of RAM usage, the most substantial variation was observed for the multi-process Nextflow, for which it increased with the number of assigned threads, while RAM consumption of the other setups did not depend much on the numbers of threads assigned for computations. Due to intermediate and log files generated, disk usage was markedly higher for the multi-process Nextflow than for the plain Bash and for the single-process Nextflow.

bioinformatics↗

bioRxiv2023

An explainable deep learning classifier of bovine mastitis based on whole genome sequence data - circumventing the p>>>n problem

The most serious drawback underlying the biological annotation of Whole Genome Sequence data is the p>>n problem, meaning that the number of polymorphic variants (p) is much larger than the number of available phenotypic records (n). Therefore, the major aim of the study was to propose a way to circumvent the problem by combining a LASSO logistic regression model with Deep Learning (DL). That was illustrated by a practical biological problem of classification of cows into mastitis-susceptible or mastitis-resistant, based on genotypes of Single Nucleotide Polymorphisms (SNPs) identified in their WGS. Among several DL architectures proposed via optimisation of DL hyperparameters using the Optuna software, imposed on different SNP sub-sets defined by LASSO logistic regressions with different penalty values, the architecture with 204,642 SNPs was selected as the best one. This architecture was composed of 2 layers with respectively 7 and 46 units per layer as well as respective drop-out rates of 0.210 and 0.358. The classification of the test data set resulted in the AUC=0.750, accuracy=0.650, sensitivity=0.600, and specificity=0.700 was selected as the best model and thus proceeded to genomic and functional annotations. Significant SNPs were selected based on the SHapley Additive exPlanation values transformed to Z-scores to assess the underlying type I-error. These SNPs were annotated to genes. As a final result, a single GO term related to the biological process and thirteen GO terms related to the molecular function were significantly enriched in the gene set that corresponded to the significant SNPs. Author SummaryOur objective is to distinguish between cows that are susceptible and resistant to bovine mastitis by analysing their genomic data. However, we face a significant challenge due to the large number of single nucleotide polymorphisms (SNPs) and limited sample size. To address this challenge, we utilize two methods: feature selection algorithms and deep learning. We experiment with various ways of implementing these techniques and evaluate their performance on a validation set. Our findings reveal that the optimal approach can accurately predict a cows susceptibility or resistance status around 65% of the time. Additionally, we employ a technique to identify the most crucial SNPs and their biological functions. Our results indicate that some of these SNPs are related to immune response or protein synthesis pathways, implying that they may affect the cows health and productivity.

bioinformatics↗

bioRxiv2022

The application of mixed linear models for the estimation of functional effects on bovine stature based on SNP summary statistics from a whole-genome association study

Genome-Wide Association Studies (GWAS) help identify polymorphic sites or genes linked to phenotypic variance, but a few identified genes / Single Nucleotide Polymorphisms are unlikely to explain a large part of the phenotypic variability of complex traits. In this study, the focus was moved from single loci to functional units, expressed by the metabolic pathways: Kyoto Encyclopaedia of Genes and Genomes (KEGG). Consequently, this study aimed to estimate KEGG effects on stature in three Nordic dairy cattle breeds using SNPs effects from GWAS as the dependent variable. The SNPs were annotated to genes, then the genes to KEGG pathways. The effects of KEGG were estimated separately for each breed using a mixed linear model incorporating the similarity between pathways expressed by common genes. The KEGG pathway D-amino acid metabolism (map00473) was estimated as significant on stature in two of the analysed breeds and revealed a borderline significance in the third breed. Interestingly, biological evidence exists that described the importance of D-amino acids for growth in experimental organisms as well as in cattle.

bioinformatics↗