A Deep-Learning Pipeline for TSS Coverage Imputation From Shallow Cell-Free DNA Sequencing.
Ontology highlight
ABSTRACT:
SUBMITTER: Han BW
PROVIDER: S-EPMC8678047 | biostudies-literature |
REPOSITORIES: biostudies-literature
Similar Datasets
Project description:Although ancient DNA data have become increasingly more important in studies about past populations, it is often not feasible or practical to obtain high coverage genomes from poorly preserved samples. While methods of accurate genotype imputation from > 1 × coverage data have recently become a routine, a large proportion of ancient samples remain unusable for downstream analyses due to their low coverage. Here, we evaluate a two-step pipeline for the imputation of common variants in ancient genomes at 0.05-1 × coverage. We use the genotype likelihood input mode in Beagle and filter for confident genotypes as the input to impute missing genotypes. This procedure, when tested on ancient genomes, outperforms a single-step imputation from genotype likelihoods, suggesting that current genotype callers do not fully account for errors in ancient sequences and additional quality controls can be beneficial. We compared the effect of various genotype likelihood calling methods, post-calling, pre-imputation and post-imputation filters, different reference panels, as well as different imputation tools. In a Neolithic Hungarian genome, we obtain ~ 90% imputation accuracy for heterozygous common variants at coverage 0.05 × and > 97% accuracy at coverage 0.5 ×. We show that imputation can mitigate, though not eliminate reference bias in ultra-low coverage ancient genomes.
| S-EPMC7596702 | biostudies-literature
Project description:Plasma DNA from 558 malignancies, 263 benign and borderline tumors and 367 healthy control samples were collected and subjected to random short-gun whole genome sequencing.
2022-07-03 | E-MTAB-10934 | biostudies-arrayexpress
Project description:Deep learning is rapidly becoming the technique of choice for automated segmentation of nuclei in biological image analysis workflows. In order to evaluate the feasibility of training nuclear segmentation models on small, custom annotated image datasets that have been augmented, we have designed a computational pipeline to systematically compare different nuclear segmentation model architectures and model training strategies. Using this approach, we demonstrate that transfer learning and tuning of training parameters, such as the composition, size, and preprocessing of the training image dataset, can lead to robust nuclear segmentation models, which match, and often exceed, the performance of existing, off-the-shelf deep learning models pretrained on large image datasets. We envision a practical scenario where deep learning nuclear segmentation models trained in this way can be shared across a laboratory, facility, or institution, and continuously improved by training them on progressively larger and varied image datasets. Our work provides computational tools and a practical framework for deep learning-based biological image segmentation using small annotated image datasets. Published [2020]. This article is a U.S. Government work and is in the public domain in the USA.
| S-EPMC8914348 | biostudies-literature
Project description:DSAP is an automated multiple-task web service designed to provide a total solution to analyzing deep-sequencing small RNA datasets generated by next-generation sequencing technology. DSAP uses a tab-delimited file as an input format, which holds the unique sequence reads (tags) and their corresponding number of copies generated by the Solexa sequencing platform. The input data will go through four analysis steps in DSAP: (i) cleanup: removal of adaptors and poly-A/T/C/G/N nucleotides; (ii) clustering: grouping of cleaned sequence tags into unique sequence clusters; (iii) non-coding RNA (ncRNA) matching: sequence homology mapping against a transcribed sequence library from the ncRNA database Rfam (http://rfam.sanger.ac.uk/); and (iv) known miRNA matching: detection of known miRNAs in miRBase (http://www.mirbase.org/) based on sequence homology. The expression levels corresponding to matched ncRNAs and miRNAs are summarized in multi-color clickable bar charts linked to external databases. DSAP is also capable of displaying miRNA expression levels from different jobs using a log(2)-scaled color matrix. Furthermore, a cross-species comparative function is also provided to show the distribution of identified miRNAs in different species as deposited in miRBase. DSAP is available at http://dsap.cgu.edu.tw.
| S-EPMC2896168 | biostudies-literature
Project description:Genotype imputation has a wide range of applications in genome-wide association study (GWAS), including increasing the statistical power of association tests, discovering trait-associated loci in meta-analyses, and prioritizing causal variants with fine-mapping. In recent years, deep learning (DL) based methods, such as sparse convolutional denoising autoencoder (SCDA), have been developed for genotype imputation. However, it remains a challenging task to optimize the learning process in DL-based methods to achieve high imputation accuracy. To address this challenge, we have developed a convolutional autoencoder (AE) model for genotype imputation and implemented a customized training loop by modifying the training process with a single batch loss rather than the average loss over batches. This modified AE imputation model was evaluated using a yeast dataset, the human leukocyte antigen (HLA) data from the 1,000 Genomes Project (1KGP), and our in-house genotype data from the Louisiana Osteoporosis Study (LOS). Our modified AE imputation model has achieved comparable or better performance than the existing SCDA model in terms of evaluation metrics such as the concordance rate (CR), the Hellinger score, the scaled Euclidean norm (SEN) score, and the imputation quality score (IQS) in all three datasets. Taking the imputation results from the HLA data as an example, the AE model achieved an average CR of 0.9468 and 0.9459, Hellinger score of 0.9765 and 0.9518, SEN score of 0.9977 and 0.9953, and IQS of 0.9515 and 0.9044 at missing ratios of 10% and 20%, respectively. As for the results of LOS data, it achieved an average CR of 0.9005, Hellinger score of 0.9384, SEN score of 0.9940, and IQS of 0.8681 at the missing ratio of 20%. In summary, our proposed method for genotype imputation has a great potential to increase the statistical power of GWAS and improve downstream post-GWAS analyses.
| S-EPMC9671213 | biostudies-literature
Project description:BackgroundRecently, a next-generation sequencing (NGS)-based method has been used for the successful detection of circulating tumor DNA (ctDNA) in various cancer types. Thus, the use of NGS on liquid biopsies will improve cancer diagnosis and prognosis. However, the low-allelic fraction of ctDNA poses a challenge for the sensitive and specific detection of tumor variants in cell-free DNA (cfDNA). To distinguish true variants from false positives, the characteristics of errors that occur during sample preparation and sequencing need to be elucidated.MethodsWe generated capture-based targeted deep sequencing data from plasma cfDNA and peripheral blood leucocyte (PBL) gDNA to profile background errors. To reveal cfDNA-associated DNA lesions, background error profiles from two sample types were compared in each nucleotide substitution class.ResultsIn this study, we determined the prevalence of single nucleotide substitutions in cfDNA sequencing data to identify DNA damage preferentially associated with cfDNA. On comparing sequencing errors between cfDNA and cellular genomic DNA (gDNA), we observed that the total substitution error rates in cfDNA were significantly higher than those in gDNA. When the substitution errors were divided into 12 substitution error classes, C:G>T:A substitution errors constituted the largest difference between cfDNA and gDNA samples. When the substitution error rates were estimated based on the location of DNA-fragment substitutions, the differences in error rates of most substitution classes between cfDNA and gDNA samples were observed only at the ends of the DNA fragments. In contrast, C:G>T:A substitution errors in the cfDNA samples were not particularly associated with DNA-fragment ends. All observations were verified in an independent dataset.ConclusionsOur data suggested that cytosine deamination increased in cfDNA compared to that in cellular gDNA. Such an observation might be due to the attenuation of DNA damage repair before the release of cfDNA and/or the accumulation of cytosine deamination after it. These findings can contribute to a better understanding of cfDNA-associated DNA damage, which will enable the accurate analysis of somatic variants present in cfDNA at an extremely low frequency.
| S-EPMC8317339 | biostudies-literature
Project description:Mutational signatures accumulate in somatic cells as an admixture of endogenous and exogenous processes that occur during an individual's lifetime. Since dividing cells release cell-free DNA (cfDNA) fragments into the circulation, we hypothesize that plasma cfDNA might reflect mutational signatures. Point mutations in plasma whole genome sequencing (WGS) are challenging to identify through conventional mutation calling due to low sequencing coverage and low mutant allele fractions. In this proof of concept study of plasma WGS at 0.3-1.5x coverage from 215 patients and 227 healthy individuals, we show that both pathological and physiological mutational signatures may be identified in plasma. By applying machine learning to mutation profiles, patients with stage I-IV cancer can be distinguished from healthy individuals with an Area Under the Curve of 0.96. Interrogating mutational processes in plasma may enable earlier cancer detection, and might enable the assessment of cancer risk and etiology.
| S-EPMC9399180 | biostudies-literature
Project description:Understanding the cell-specific binding patterns of transcription factors (TFs) is fundamental to studying gene regulatory networks in biological systems, for which ChIP-seq not only provides valuable data but is also considered as the gold standard. Despite tremendous efforts from the scientific community to conduct TF ChIP-seq experiments, the available data represent only a limited percentage of ChIP-seq experiments, considering all possible combinations of TFs and cell lines. In this study, we demonstrate a method for accurately predicting cell-specific TF binding for TF-cell line combinations based on only a small fraction (4%) of the combinations using available ChIP-seq data. The proposed model, termed TFImpute, is based on a deep neural network with a multi-task learning setting to borrow information across transcription factors and cell lines. Compared with existing methods, TFImpute achieves comparable accuracy on TF-cell line combinations with ChIP-seq data; moreover, TFImpute achieves better accuracy on TF-cell line combinations without ChIP-seq data. This approach can predict cell line specific enhancer activities in K562 and HepG2 cell lines, as measured by massively parallel reporter assays, and predicts the impact of SNPs on TF binding.
| S-EPMC5345877 | biostudies-literature
Project description:A question of fundamental biological significance is to what extent the expression of a subset of genes can be used to recover the full transcriptome, with important implications for biological discovery and clinical application. To address this challenge, we propose two novel deep learning methods, PMI and GAIN-GTEx, for gene expression imputation. In order to increase the applicability of our approach, we leverage data from GTEx v8, a reference resource that has generated a comprehensive collection of transcriptomes from a diverse set of human tissues. We show that our approaches compare favorably to several standard and state-of-the-art imputation methods in terms of predictive performance and runtime in two case studies and two imputation scenarios. In comparison conducted on the protein-coding genes, PMI attains the highest performance in inductive imputation whereas GAIN-GTEx outperforms the other methods in in-place imputation. Furthermore, our results indicate strong generalization on RNA-Seq data from 3 cancer types across varying levels of missingness. Our work can facilitate a cost-effective integration of large-scale RNA biorepositories into genomic studies of disease, with high applicability across diverse tissue types.
| S-EPMC8076954 | biostudies-literature
Project description:Aerosols have adverse health effects and play a significant role in the climate as well. The Multiangle Implementation of Atmospheric Correction (MAIAC) provides Aerosol Optical Depth (AOD) at high temporal (daily) and spatial (1 km) resolution, making it particularly useful to infer and characterize spatiotemporal variability of aerosols at a fine spatial scale for exposure assessment and health studies. However, clouds and conditions of high surface reflectance result in a significant proportion of missing MAIAC AOD. To fill these gaps, we present an imputation approach using deep learning with downscaling. Using a baseline autoencoder, we leverage residual connections in deep neural networks to boost learning and parameter sharing to reduce overfitting, and conduct bagging to reduce error variance in the imputations. Downscaled through a similar auto-encoder based deep residual network, Modern-Era Retrospective analysis for Research and Applications Version 2 (MERRA-2) GMI Replay Simulation (M2GMI) data were introduced to the network as an important gap-filling feature that varies in space to be used for missingness imputations. Imputing weekly MAIAC AOD from 2000 to 2016 over California, a state with considerable geographic heterogeneity, our full (non-full) residual network achieved mean R2 = 0.94 (0.86) [RMSE = 0.007 (0.01)] in an independent test, showing considerably better performance than a regular neural network or non-linear generalized additive model (mean R2 = 0.78-0.81; mean RMSE = 0.013-0.015). The adjusted imputed as well as combined imputed and observed MAIAC AOD showed strong correlation with Aerosol Robotic Network (AERONET) AOD (R = 0.83; R2 = 0.69, RMSE = 0.04). Our results show that we can generate reliable imputations of missing AOD through a deep learning approach, having important downstream air quality modeling applications.
| S-EPMC7063693 | biostudies-literature