Dataset Information

Recapitulation of patient-specific 3D chromatin conformation using machine learning.

ABSTRACT: Regulatory networks containing enhancer-gene edges define cellular states. Multiple efforts have revealed these networks for reference tissues and cell lines by integrating multi-omics data. However, the methods developed cannot be applied for large patient cohorts due to the infeasibility of chromatin immunoprecipitation sequencing (ChIP-seq) for limited biopsy material. We trained machine-learning models using chromatin interaction analysis with paired-end tag sequencing (ChIA-PET) and high-throughput chromosome conformation capture combined with chromatin immunoprecipitation (HiChIP) data that can predict connections using only assay for transposase-accessible chromatin using sequencing (ATAC-seq) and RNA-seq data as input, which can be generated from biopsies. Our method overcomes limitations of correlation-based approaches that cannot distinguish between distinct target genes of given enhancers or between active vs. poised states in different samples, a hallmark of network rewiring in cancer. Application of our model on 371 samples across 22 cancer types revealed 1,780 enhancer-gene connections for 602 cancer genes. Using CRISPR interference (CRISPRi), we validated enhancers predicted to regulate ESR1 in estrogen receptor (ER)+ breast cancer and A1CF in liver hepatocellular carcinoma.

SUBMITTER: Xu D

PROVIDER: S-EPMC10545938 | biostudies-literature | 2023 Sep

REPOSITORIES: biostudies-literature

ACCESS DATA

Publications

Recapitulation of patient-specific 3D chromatin conformation using machine learning.

Xu Duo D Forbes Andre Neil AN Cohen Sandra S Palladino Ann A Karadimitriou Tatiana T Khurana Ekta E

Cell reports methods 20230905 9

Regulatory networks containing enhancer-gene edges define cellular states. Multiple efforts have revealed these networks for reference tissues and cell lines by integrating multi-omics data. However, the methods developed cannot be applied for large patient cohorts due to the infeasibility of chromatin immunoprecipitation sequencing (ChIP-seq) for limited biopsy material. We trained machine-learning models using chromatin interaction analysis with paired-end tag sequencing (ChIA-PET) and high-th ...[more]

PMID: 37673071

Similar Datasets

Project description:BACKGROUND AND PURPOSE:MR imaging-based modeling of tumor cell density can substantially improve targeted treatment of glioblastoma. Unfortunately, interpatient variability limits the predictive ability of many modeling approaches. We present a transfer learning method that generates individualized patient models, grounded in the wealth of population data, while also detecting and adjusting for interpatient variabilities based on each patient's own histologic data. MATERIALS AND METHODS:We recruited patients with primary glioblastoma undergoing image-guided biopsies and preoperative imaging, including contrast-enhanced MR imaging, dynamic susceptibility contrast MR imaging, and diffusion tensor imaging. We calculated relative cerebral blood volume from DSC-MR imaging and mean diffusivity and fractional anisotropy from DTI. Following image coregistration, we assessed tumor cell density for each biopsy and identified corresponding localized MR imaging measurements. We then explored a range of univariate and multivariate predictive models of tumor cell density based on MR imaging measurements in a generalized one-model-fits-all approach. We then implemented both univariate and multivariate individualized transfer learning predictive models, which harness the available population-level data but allow individual variability in their predictions. Finally, we compared Pearson correlation coefficients and mean absolute error between the individualized transfer learning and generalized one-model-fits-all models. RESULTS:Tumor cell density significantly correlated with relative CBV (r = 0.33, P < .001), and T1-weighted postcontrast (r = 0.36, P < .001) on univariate analysis after correcting for multiple comparisons. With single-variable modeling (using relative CBV), transfer learning increased predictive performance (r = 0.53, mean absolute error = 15.19%) compared with one-model-fits-all (r = 0.27, mean absolute error = 17.79%). With multivariate modeling, transfer learning further improved performance (r = 0.88, mean absolute error = 5.66%) compared with one-model-fits-all (r = 0.39, mean absolute error = 16.55%). CONCLUSIONS:Transfer learning significantly improves predictive modeling performance for quantifying tumor cell density in glioblastoma.

Project description:Introduction: The air puff test is a contactless tonometry test used to measure the intraocular pressure and the cornea's biomechanical properties. Limitations that most challenge the accuracy of the estimation of the corneal material and the intraocular pressure are the strong intercorrelation between the intraocular pressure and the corneal parameters, either the material properties that can change from one person to another because of age or the geometry parameters like central corneal thickness. This influence produces inaccuracies in the corneal deformation parameters while extracting the IOP parametric equation, which can be reduced through the consideration of the patient-specific air puff pressure distribution taking into account the changes in corneal parameters. This air puff pressure loading distribution can be determined precisely from the fluid-structure interaction (FSI) coupling between the air puff and the eye model. However, the computational fluid dynamics simulation of the air puff in the coupling algorithm is a time-consuming model that is impractical to use in clinical practice and large parametric studies. Methods: By using a supervised machine learning algorithm, we predict the time-dependent air puff pressure distribution for different corneal parameters via a parametric study of the corneal deformations and the gradient boosting algorithm. Results: The results confirmed that the algorithm gives the time-dependent air puff pressure distribution with an MAE of 0.0258, an RMSE of 0.0673, and an execution time of 93 s, which is then applied to the finite element model of the eye generating the corresponding corneal deformations taking into account the FSI influence. Using corneal deformations, the response parameters can be extracted and used to produce more accurate algorithms of the intraocular pressure and corneal material stress-strain index (SSI). Discussion: Estimating the distribution of air pressure on the cornea is essential to increase the accuracy of intraocular pressure (IOP) measurements, which serve as valuable indicator of corneal disease. We find that the air puff pressure loading is largely influenced by complex changes in corneal parameters unique to each patient case. With our innovative algorithm, we can preserve the same accuracy developed by the CFD-based FSI model, while reducing the computational time from approximately 101000 s (28 h) to 720 s (12 min), which is about 99.2% reduction in time. This huge improvement in computational cost will lead to significant improvement in the parametric equations for IOP and the Stress-Strain Index (SSI).

Project description:Digital phantoms are important tools for optimization and evaluation of x-ray imaging systems, and should ideally reflect the 3D structure of human anatomy and its potential variability. In addition, they need to include a good level of detail at a high enough spatial resolution to accurately model the continuous nature of the human anatomy. A pipeline to increase the spatial resolution of patient-based digital breast phantoms that can be used for computer simulations of breast imaging is proposed. Given a tomographic breast image of finite resolution, the proposed methods can generate a phantom and increase its resolution at will, not only simply through super-sampling, but also by generating additional random glandular details to account for glandular edges and strands to compensate for those that may have not been detected in the original image due to the limited spatial resolution of the imaging system used. The proposed algorithms use supervised learning to predict the loss in glandularity due to limited resolution, and then to realistically recover this loss by learning the mapping between low and high resolution images. They were trained on high-resolution synchrotron images (detector pixel size 60 ?m) reconstructed at seven voxel dimensions (60 ?m-480 ?m), and applied to patient images acquired with a clinical breast CT system (detector pixel size 194 ?m) to generate super-resolution phantoms (voxel sizes 68 ?m). Several evaluations were made to assess the appropriateness of the developed methods, both with the synchrotron (relative prediction error 0.010??±??0.004, recovering accuracy 0.95??±??0.04), and with the clinical images (average glandularity error at 194 ?m: 0.15%??±??0.12%). Finally, a breast radiologist assessed the realism of the developed phantoms by blindly comparing original and phantom images, resulting in not being able to distinguish the real from the phantom images. In conclusion, the proposed method can generate super-resolution phantoms from tomographic breast patient images that can be used for future computer simulations for optimization of new breast imaging technologies.

Dataset Information

Recapitulation of patient-specific 3D chromatin conformation using machine learning.

Publications

Recapitulation of patient-specific 3D chromatin conformation using machine learning.

Similar Datasets

OmicsDI is part of the ELIXIR infrastructure

Tweets