PMCID string | Title string | Sentences string |
|---|---|---|
PMC11481779 | HiCDiffusion - diffusion-enhanced, transformer-based prediction of chromatin interactions from DNA sequences | The input channels of the residual blocks used in the encoder are: (5, 32, 32, 32, 64, 64, 64, 128, 128, 256, 256, 256, 256), and the output channels are: (32, 32, 32, 64, 64, 64, 128, 128, 256, 256, 256, 256, 256). |
PMC11481779 | HiCDiffusion - diffusion-enhanced, transformer-based prediction of chromatin interactions from DNA sequences | The encoder has one additional layer - the transformer encoder - that helps to learn the appropriate context of the latent representation produced by convolutions. |
PMC11481779 | HiCDiffusion - diffusion-enhanced, transformer-based prediction of chromatin interactions from DNA sequences | The transformer encoder used in our study uses 8 heads. |
PMC11481779 | HiCDiffusion - diffusion-enhanced, transformer-based prediction of chromatin interactions from DNA sequences | However, we also use residual connections around the transformer to preserve original data and help with the vanishing gradient problem. |
PMC11481779 | HiCDiffusion - diffusion-enhanced, transformer-based prediction of chromatin interactions from DNA sequences | Finally, we apply ReLU to the output and expand the result to the 2D matrix with 512 dimensions - by repeating the (256, 256) output vertically and horizontally. |
PMC11481779 | HiCDiffusion - diffusion-enhanced, transformer-based prediction of chromatin interactions from DNA sequences | That procedure leaves us with two outputs of (256, 256, 256) - one for repeating vertically and one for horizontally. |
PMC11481779 | HiCDiffusion - diffusion-enhanced, transformer-based prediction of chromatin interactions from DNA sequences | We further concatenate it to the final output of the encoder of (512, 256, 256). |
PMC11481779 | HiCDiffusion - diffusion-enhanced, transformer-based prediction of chromatin interactions from DNA sequences | The decoder architecture is composed of 2D residual convolutions - each of which is composed of 2D convolution (with 512 input and output channels, kernel size of 3, and padding of 1), batch normalisation, ReLU function, and again exactly the same 2D convolution, and batch normalisation. |
PMC11481779 | HiCDiffusion - diffusion-enhanced, transformer-based prediction of chromatin interactions from DNA sequences | The output of the 2D residual block is created by applying the ReLU function to the sum of the output of the previously explained sequence and the input to the residual block - this time, unlikely in the Encoder, without any downscaling. |
PMC11481779 | HiCDiffusion - diffusion-enhanced, transformer-based prediction of chromatin interactions from DNA sequences | It is crucial in the decoder to use various options for the dilation parameter - it has been done in the previously mentioned papers. |
PMC11481779 | HiCDiffusion - diffusion-enhanced, transformer-based prediction of chromatin interactions from DNA sequences | It is also done in our work to propagate information through the whole heatmap. |
PMC11481779 | HiCDiffusion - diffusion-enhanced, transformer-based prediction of chromatin interactions from DNA sequences | For each next layer, the dilation parameter is set to (2, 4, 8, 16, 32) and represents exponential growth. |
PMC11481779 | HiCDiffusion - diffusion-enhanced, transformer-based prediction of chromatin interactions from DNA sequences | Finally, after the decoder is done, we apply an additional convolutional layer (kernel size of 3 and padding of 1) that changes dimensions from 512 to 1, leaving us with the heatmap of size (256, 256). |
PMC11481779 | HiCDiffusion - diffusion-enhanced, transformer-based prediction of chromatin interactions from DNA sequences | We apply L1 loss function to match the original heatmap. |
PMC11481779 | HiCDiffusion - diffusion-enhanced, transformer-based prediction of chromatin interactions from DNA sequences | The previously explained encoder-decoder architecture is then used for further processing by using transfer learning - we pre-train the encoder-decoder and use it for the final architecture based on diffusion models. |
PMC11481779 | HiCDiffusion - diffusion-enhanced, transformer-based prediction of chromatin interactions from DNA sequences | The final architecture first computes the latent representation (output of the encoder & decoder - with 512 channel), and the final heatmap from the encoder-decoder architecture. |
PMC11481779 | HiCDiffusion - diffusion-enhanced, transformer-based prediction of chromatin interactions from DNA sequences | We compute the residual heatmap by taking the difference between the real heatmap and the one produced by the encoder-decoder architecture. |
PMC11481779 | HiCDiffusion - diffusion-enhanced, transformer-based prediction of chromatin interactions from DNA sequences | The network’s target is predicting the residual heatmap, which shows us how to improve the prediction of the encoder-decoder architecture. |
PMC11481779 | HiCDiffusion - diffusion-enhanced, transformer-based prediction of chromatin interactions from DNA sequences | The diffusion models work in two steps - the first is the model’s training, and the second is inference. |
PMC11481779 | HiCDiffusion - diffusion-enhanced, transformer-based prediction of chromatin interactions from DNA sequences | The training comes first so that it will be explained first as well. |
PMC11481779 | HiCDiffusion - diffusion-enhanced, transformer-based prediction of chromatin interactions from DNA sequences | We take the residual heatmap and apply Gaussian noise multiple times - up to T. Then, we take such a noised image and use a denoising U-Net network. |
PMC11481779 | HiCDiffusion - diffusion-enhanced, transformer-based prediction of chromatin interactions from DNA sequences | We try to predict the noise added to the image. |
PMC11481779 | HiCDiffusion - diffusion-enhanced, transformer-based prediction of chromatin interactions from DNA sequences | The implementation of the diffusion model is pytorch reimplementation of the original paper . |
PMC11481779 | HiCDiffusion - diffusion-enhanced, transformer-based prediction of chromatin interactions from DNA sequences | We have made a few changes to make it work with the Hi-C data. |
PMC11481779 | HiCDiffusion - diffusion-enhanced, transformer-based prediction of chromatin interactions from DNA sequences | Firstly, the U-Net is conditioned by taking latent encoder-decoder output (512, 256, 256). |
PMC11481779 | HiCDiffusion - diffusion-enhanced, transformer-based prediction of chromatin interactions from DNA sequences | It is then converted in the initial phase of using U-Net to (32, 256, 256) using convolution with a kernel equal to 7, and padding of 3. |
PMC11481779 | HiCDiffusion - diffusion-enhanced, transformer-based prediction of chromatin interactions from DNA sequences | The original image provided to U-Net (noised image) is also processed using convolution to sizes of (32, 256, 256). |
PMC11481779 | HiCDiffusion - diffusion-enhanced, transformer-based prediction of chromatin interactions from DNA sequences | The condition and noised image upscaled to 32 channels are then concatenated and fed into the network to represent the conditional diffusion process. |
PMC11481779 | HiCDiffusion - diffusion-enhanced, transformer-based prediction of chromatin interactions from DNA sequences | We use then MSE loss function to match with the desired output. |
PMC11481779 | HiCDiffusion - diffusion-enhanced, transformer-based prediction of chromatin interactions from DNA sequences | After training the diffusion model, we can predict the residual Hi-C matrix using inference mode - that is, in the form of guided diffusion using encoder-decoder output from random noise, which allows us to obtain more realistically looking Hi-C matrices. |
PMC11481779 | HiCDiffusion - diffusion-enhanced, transformer-based prediction of chromatin interactions from DNA sequences | All the models were tested using independent chromosomes not used in the training or validation procedure. |
PMC11481779 | HiCDiffusion - diffusion-enhanced, transformer-based prediction of chromatin interactions from DNA sequences | We have calculated the Pearson correlation coefficient and SCC for each example and Fréchet inception distance to compare the quality of the images. |
PMC11481779 | HiCDiffusion - diffusion-enhanced, transformer-based prediction of chromatin interactions from DNA sequences | We have run a version of C.Origami that includes only sequence (without CTCF and ATAC-Seq signals), and compared Pearson correlation coefficient scores, SCCs, and FID scores. |
PMC11481779 | HiCDiffusion - diffusion-enhanced, transformer-based prediction of chromatin interactions from DNA sequences | Fréchet inception distance is a metric used most often in computer vision to measure the quality of the images. |
PMC11481779 | HiCDiffusion - diffusion-enhanced, transformer-based prediction of chromatin interactions from DNA sequences | It is formally defined as:[12pt] $$\:FID\:=\:(__}^\:}^}^||x-y|^d\:(x,y))}^}$$ Where is the set of all measures on with as first marginal factor, and as the second. |
PMC11481779 | HiCDiffusion - diffusion-enhanced, transformer-based prediction of chromatin interactions from DNA sequences | In our work, we are using torchmetrics , which solves the aforementioned equation for multidimensional Gaussian distributions as:[12pt] $$\:FID=||\:-_|^+tr(_-2(\:_^}}_)$$ In which is the mean, and is the variance of a multivariate normal distribution estimated from Inception v3 features calculated on real data (true Hi-C matrices) and and are corresponding values for the multivariate normal distribution estimated on the generated data (predicted Hi-C matrices). |
PMC11481779 | HiCDiffusion - diffusion-enhanced, transformer-based prediction of chromatin interactions from DNA sequences | Since this metric is taking images, we need to normalise our output to the range and add 2 additional channels - which is done by repeating the data. |
PMC11481779 | HiCDiffusion - diffusion-enhanced, transformer-based prediction of chromatin interactions from DNA sequences | The output of the procedure is called the FID score, and the lower it is, the higher quality the image is - as it is comparing the distributions of the generated images to the real ones (or, in our case, Hi-C matrices). |
PMC11481779 | HiCDiffusion - diffusion-enhanced, transformer-based prediction of chromatin interactions from DNA sequences | To compare the results to the C.Origami, the current state-of-the-art tool for predicting 3D interactions, we have downloaded the original software with the dataset. |
PMC11481779 | HiCDiffusion - diffusion-enhanced, transformer-based prediction of chromatin interactions from DNA sequences | However, we have used two comparison models since our study focuses on pure sequence-to-interaction relationships. |
PMC11481779 | HiCDiffusion - diffusion-enhanced, transformer-based prediction of chromatin interactions from DNA sequences | In the first one, we modified the tool to accept DNA sequence and train only on it. |
PMC11481779 | HiCDiffusion - diffusion-enhanced, transformer-based prediction of chromatin interactions from DNA sequences | The second one was the full C.Origami architecture - the one that includes CTCF and ATAC-Seq epigenetic signals. |
PMC11481779 | HiCDiffusion - diffusion-enhanced, transformer-based prediction of chromatin interactions from DNA sequences | Then, we trained both C.Origami models, with validation chromosome chr10, and testing chromosome chr15. |
PMC11481779 | HiCDiffusion - diffusion-enhanced, transformer-based prediction of chromatin interactions from DNA sequences | This approach allowed us to get accurate metrics for testing and validation chromosomes and the supremum of the metrics in the case of training chromosomes. |
PMC11481779 | HiCDiffusion - diffusion-enhanced, transformer-based prediction of chromatin interactions from DNA sequences | Furthermore, this model was used to predict precisely the same windows as in our models. |
PMC11481779 | HiCDiffusion - diffusion-enhanced, transformer-based prediction of chromatin interactions from DNA sequences | We have calculated the Pearson correlation coefficient and SCCs for all the predictions and FID scores for all chromosomes. |
PMC11481779 | HiCDiffusion - diffusion-enhanced, transformer-based prediction of chromatin interactions from DNA sequences | The Pearson correlation coefficient, while used in many studies, is often criticised in the case of comparisons between Hi-C matrices. |
PMC11481779 | HiCDiffusion - diffusion-enhanced, transformer-based prediction of chromatin interactions from DNA sequences | We have decided to use a stratum-adjusted correlation coefficient (SCC) to show the usefulness of our algorithm. |
PMC11481779 | HiCDiffusion - diffusion-enhanced, transformer-based prediction of chromatin interactions from DNA sequences | We have used Python implementation of the algorithm . |
PMC11481779 | HiCDiffusion - diffusion-enhanced, transformer-based prediction of chromatin interactions from DNA sequences | Since our matrices are in the form of 256 × 256, we have used h = 2 for smoothening and the maximum distance of 16 (equivalent to more than 100kbp). |
PMC11481779 | HiCDiffusion - diffusion-enhanced, transformer-based prediction of chromatin interactions from DNA sequences | To call TADs, we used the engine of the TADBit . |
PMC11481779 | HiCDiffusion - diffusion-enhanced, transformer-based prediction of chromatin interactions from DNA sequences | We have implemented loading the numpy / tensor arrays into their data structures, as the original software takes as the input. |
PMC11481779 | HiCDiffusion - diffusion-enhanced, transformer-based prediction of chromatin interactions from DNA sequences | cool files. |
PMC11481779 | HiCDiffusion - diffusion-enhanced, transformer-based prediction of chromatin interactions from DNA sequences | We have then calculated the insulation score for each of the heatmaps independently, and transformed it into TAD borders. |
PMC11481779 | HiCDiffusion - diffusion-enhanced, transformer-based prediction of chromatin interactions from DNA sequences | We have calculated pearson correlation of the insulation scores for all heatmaps (we have filtered out the ones that in real Hi-C data are starting/ending with zeros, as well as the ones that contain more than 5% of zero values). |
PMC11481779 | HiCDiffusion - diffusion-enhanced, transformer-based prediction of chromatin interactions from DNA sequences | To show that, in principle the loop calling works on the modelled data, we have implemented a proof-of-concept algorithm similar to HiCCUPS/BH-FDR algorithm. |
PMC11481779 | HiCDiffusion - diffusion-enhanced, transformer-based prediction of chromatin interactions from DNA sequences | However, in our case we only considered the four regions (donut, top-bottom, left-right, and left-bottom), without applying any corrections for ease of implementation, and to show that the loop calling is actually possible on the predicted data. |
PMC11481779 | HiCDiffusion - diffusion-enhanced, transformer-based prediction of chromatin interactions from DNA sequences | Additionally, since HiCDiffusion does not use any epigenetic information, it is at strong disadvantage in comparison to methods that get CTCF-related information, like epigenetic tracks, or just regular peak callings. |
PMC11481779 | HiCDiffusion - diffusion-enhanced, transformer-based prediction of chromatin interactions from DNA sequences | To simulate transduplication, a model trained on all chromosomes except chromosome 8 was taken into account. |
PMC11481779 | HiCDiffusion - diffusion-enhanced, transformer-based prediction of chromatin interactions from DNA sequences | To illustrate that the model can correctly predict the effects of such a mutation, chr8:32 100 000–34 100 000 region was taken as an example. |
PMC11481779 | HiCDiffusion - diffusion-enhanced, transformer-based prediction of chromatin interactions from DNA sequences | We have then copied the sequence of 200kbp, precisely chr8:33 450 000–33 650 000, and inserted it instead of sequence present at chr8:32 750 000–32 950 000. |
PMC11481779 | HiCDiffusion - diffusion-enhanced, transformer-based prediction of chromatin interactions from DNA sequences | The model was then used to predict the in-sillico effect of such trans-duplication. |
PMC11481779 | HiCDiffusion - diffusion-enhanced, transformer-based prediction of chromatin interactions from DNA sequences | To test the human-derived model on the mouse data, we took the model trained on all chromosomes except chromosome 8 and tested it among all chromosomes of the mouse (as it is a completely different organism, the training/validation data is independent of the mouse). |
PMC11481779 | HiCDiffusion - diffusion-enhanced, transformer-based prediction of chromatin interactions from DNA sequences | We used the mm10 reference genome and excluded centromeres and telomeres from the analysis. |
PMC11481779 | HiCDiffusion - diffusion-enhanced, transformer-based prediction of chromatin interactions from DNA sequences | As the human model is trained on GM12878 data (which is lymphoblastoid cells), we have decided to use a similar cell line in mouse. |
PMC11481779 | HiCDiffusion - diffusion-enhanced, transformer-based prediction of chromatin interactions from DNA sequences | We have taken B cell derived cell line from the 4DN database (mcool file accession number: 4DNFIPNP9H9T, experiment accession number: 4DNESYX7AQRY). |
PMC11481779 | HiCDiffusion - diffusion-enhanced, transformer-based prediction of chromatin interactions from DNA sequences | We have calculated the correlation scores similarly as in the case of the human data. |
PMC11481779 | HiCDiffusion - diffusion-enhanced, transformer-based prediction of chromatin interactions from DNA sequences | In our study, we have created a novel deep learning architecture, HiCDiffusion, that is based on the latest advances in the field of computer vision and Artificial Intelligence. |
PMC11481779 | HiCDiffusion - diffusion-enhanced, transformer-based prediction of chromatin interactions from DNA sequences | The model, composed of an encoder, transformer encoder, decoder, and a diffusion network, has surpassed the quality metric - FID score on average by 12 times (sequence-only comparison; by 11 times in case of C.Origami with epigenetics), while in the best-case chromosome, improvement was 88 times (sequence-only comparison as well; 56 in case of epigenetics-enhanced C.Origami version). |
PMC11481779 | HiCDiffusion - diffusion-enhanced, transformer-based prediction of chromatin interactions from DNA sequences | The level of the artificiality of the in silico HiC images (true quality) can also be seen with the human eye. |
PMC11481779 | HiCDiffusion - diffusion-enhanced, transformer-based prediction of chromatin interactions from DNA sequences | In the case of the current models, like the aforementioned C.Origami or Akita, the model output is blurred. |
PMC11481779 | HiCDiffusion - diffusion-enhanced, transformer-based prediction of chromatin interactions from DNA sequences | It can be easily distinguished from real data. |
PMC11481779 | HiCDiffusion - diffusion-enhanced, transformer-based prediction of chromatin interactions from DNA sequences | In the case of our model, we have obtained quality that can be easily confused with the experimental data while obtaining all the metrics that have made previously proposed models great. |
PMC11481779 | HiCDiffusion - diffusion-enhanced, transformer-based prediction of chromatin interactions from DNA sequences | Our method is the next step towards obtaining a reliable and functional universal predictor of the spatial organisation of the chromatin within the nucleus, using purely DNA sequence - which would make connecting population studies with 3D genomics much easier and, foremost - cheaper. |
PMC9430060 | P1289: COMPUTATIONAL SIMULATIONS SHED LIGHT ON THE 3D CONSEQUENCE OF SUPER-ENHANCER HIJACKING IN B CELL MALIGNANCIES | Background: Chromosomal translocations are important drivers of haematological malignancies whereby proto-oncogenes are activated by juxtaposition with super-enhancers. |
PMC9430060 | P1289: COMPUTATIONAL SIMULATIONS SHED LIGHT ON THE 3D CONSEQUENCE OF SUPER-ENHANCER HIJACKING IN B CELL MALIGNANCIES | The immunoglobulin (IG) and T-cell receptor loci are enhancer-rich regions that are commonly hijacked via translocation in many subtypes of lymphoid malignancies. |
PMC9430060 | P1289: COMPUTATIONAL SIMULATIONS SHED LIGHT ON THE 3D CONSEQUENCE OF SUPER-ENHANCER HIJACKING IN B CELL MALIGNANCIES | However, the epigenomic landscape and 3D genome structure remains poorly understood. |
PMC9430060 | P1289: COMPUTATIONAL SIMULATIONS SHED LIGHT ON THE 3D CONSEQUENCE OF SUPER-ENHANCER HIJACKING IN B CELL MALIGNANCIES | We previously analysed the epigenomic consequence of rearrangements between the super-enhancers of the IG heavy chain locus (IGH) and proto-oncogene CCND1. |
PMC9430060 | P1289: COMPUTATIONAL SIMULATIONS SHED LIGHT ON THE 3D CONSEQUENCE OF SUPER-ENHANCER HIJACKING IN B CELL MALIGNANCIES | By integrating BLUEPRINT epigenomic data with DNA breakpoint detection, we observed a broad H3K4me3 domain (BD) within the IGH locus of healthy B-cells that was absent in samples with IGH-CCND1 translocations. |
PMC9430060 | P1289: COMPUTATIONAL SIMULATIONS SHED LIGHT ON THE 3D CONSEQUENCE OF SUPER-ENHANCER HIJACKING IN B CELL MALIGNANCIES | Upon chromosomal translocation, we observed the appearance of an H3K4me3-BD over CCND1 concomitant with its overexpression and an extensive increase in chromatin accessibility of this locus. |
PMC9430060 | P1289: COMPUTATIONAL SIMULATIONS SHED LIGHT ON THE 3D CONSEQUENCE OF SUPER-ENHANCER HIJACKING IN B CELL MALIGNANCIES | In addition, we observed similar cancer-specific H3K4me3-BDs associated with super-enhancer hijacking of other common oncogenes in B-cell (MAF, MYC and FGFR3/NSD2) and T-cell malignancies (LMO2, TLX3 and TAL1). |
PMC9430060 | P1289: COMPUTATIONAL SIMULATIONS SHED LIGHT ON THE 3D CONSEQUENCE OF SUPER-ENHANCER HIJACKING IN B CELL MALIGNANCIES | Aims: To understand and predict the 3D genome consequence of IGH-CCND1 translocation events using the highly predictive heteromorphic polymer model (HiP-HoP). |
PMC9430060 | P1289: COMPUTATIONAL SIMULATIONS SHED LIGHT ON THE 3D CONSEQUENCE OF SUPER-ENHANCER HIJACKING IN B CELL MALIGNANCIES | Methods: Using methods from polymer physics, HiP-HoP uses data on DNA accessibility, chromatin states, and protein binding to generate an ensemble of simulated 3D chromosome structures. |
PMC9430060 | P1289: COMPUTATIONAL SIMULATIONS SHED LIGHT ON THE 3D CONSEQUENCE OF SUPER-ENHANCER HIJACKING IN B CELL MALIGNANCIES | After confirming that the model gives good predictions of Hi-C data for a healthy human B-cell derived cell line, we generated predictions for two B-cell malignant cell lines (U266, Z-138) possessing IGH-CCND1 rearrangements. |
PMC9430060 | P1289: COMPUTATIONAL SIMULATIONS SHED LIGHT ON THE 3D CONSEQUENCE OF SUPER-ENHANCER HIJACKING IN B CELL MALIGNANCIES | Results: We confirmed that the insertion of a single IGH super enhancer (Eα1) in U266 leaves the local topologically associated domain (TAD) structure intact. |
PMC9430060 | P1289: COMPUTATIONAL SIMULATIONS SHED LIGHT ON THE 3D CONSEQUENCE OF SUPER-ENHANCER HIJACKING IN B CELL MALIGNANCIES | Simulations of the reciprocal IGH-CCND1 translocation in Z-138, suggested the TAD boundaries are preserved, creating an oncogenic fusion TAD encompassing CCND1 and all four IGH super-enhancers. |
PMC9430060 | P1289: COMPUTATIONAL SIMULATIONS SHED LIGHT ON THE 3D CONSEQUENCE OF SUPER-ENHANCER HIJACKING IN B CELL MALIGNANCIES | Further simulations suggested that it is the downstream epigenetic changes which drive the 3D structural changes, rather than the genomic rearrangement itself. |
PMC9430060 | P1289: COMPUTATIONAL SIMULATIONS SHED LIGHT ON THE 3D CONSEQUENCE OF SUPER-ENHANCER HIJACKING IN B CELL MALIGNANCIES | This points to a scenario where the translocated IGH super-enhancers first recruit chromatin remodelers to the region; this then drives epigenomic changes which in turn disrupts the wider 3D structure leading to dysregulation of cyclin D1. |
PMC9430060 | P1289: COMPUTATIONAL SIMULATIONS SHED LIGHT ON THE 3D CONSEQUENCE OF SUPER-ENHANCER HIJACKING IN B CELL MALIGNANCIES | The model was able to predict how the structure of CCND1 changes in these different cell lines, could simulate different breakpoints, and could predict the impact on gene expression. |
PMC9430060 | P1289: COMPUTATIONAL SIMULATIONS SHED LIGHT ON THE 3D CONSEQUENCE OF SUPER-ENHANCER HIJACKING IN B CELL MALIGNANCIES | Using qPCR and 3D-FISH we were able to validate the predicted high mRNA expression levels and increase in volume of the CCND1 locus. |
PMC9430060 | P1289: COMPUTATIONAL SIMULATIONS SHED LIGHT ON THE 3D CONSEQUENCE OF SUPER-ENHANCER HIJACKING IN B CELL MALIGNANCIES | Summary/Conclusion: Taken together, our analysis suggests that H3K4me3-BDs are created by super-enhancers and supports the new concept of epigenomic translocation, where the relocation of H3K4me3-BDs accompanies the translocation of super-enhancers, altering the local 3D genome. |
PMC9430060 | P1289: COMPUTATIONAL SIMULATIONS SHED LIGHT ON THE 3D CONSEQUENCE OF SUPER-ENHANCER HIJACKING IN B CELL MALIGNANCIES | References 1. |
PMC9430060 | P1289: COMPUTATIONAL SIMULATIONS SHED LIGHT ON THE 3D CONSEQUENCE OF SUPER-ENHANCER HIJACKING IN B CELL MALIGNANCIES | PMID:34933939 |
PMC11225860 | Aurkin-A, a TPX2-aurora a small molecule inhibitor disrupts Alisertib-induced polyploidy in aggressive diffuse large B cell lymphoma | •Alisertib, an ATP site inhibitor of Aurora A kinase, induces cell cycle dysregulation and polyploidy in Diffuse Large B Cell Lymphoma.•The combination of alisertib and Aurkin A, an Aurora A TPX2 site inhibitor, prevents polyploidy and increases apoptosis. |
PMC11225860 | Aurkin-A, a TPX2-aurora a small molecule inhibitor disrupts Alisertib-induced polyploidy in aggressive diffuse large B cell lymphoma | Alisertib, an ATP site inhibitor of Aurora A kinase, induces cell cycle dysregulation and polyploidy in Diffuse Large B Cell Lymphoma. |
PMC11225860 | Aurkin-A, a TPX2-aurora a small molecule inhibitor disrupts Alisertib-induced polyploidy in aggressive diffuse large B cell lymphoma | The combination of alisertib and Aurkin A, an Aurora A TPX2 site inhibitor, prevents polyploidy and increases apoptosis. |
PMC11225860 | Aurkin-A, a TPX2-aurora a small molecule inhibitor disrupts Alisertib-induced polyploidy in aggressive diffuse large B cell lymphoma | Chromosome instability (CIN) and aneuploidy are classical hallmarks of cancer, caused by gross genomic rearrangements and cell cycle slippage . |
PMC11225860 | Aurkin-A, a TPX2-aurora a small molecule inhibitor disrupts Alisertib-induced polyploidy in aggressive diffuse large B cell lymphoma | In a tumor, CIN is variable from cell to cell and can include changes in chromosome structure, copy number variations (gains and deep deletions), translocation, and chromosomal rearrangements [, , , ]. |
PMC11225860 | Aurkin-A, a TPX2-aurora a small molecule inhibitor disrupts Alisertib-induced polyploidy in aggressive diffuse large B cell lymphoma | Aneuploidy is a direct outcome of this instability. |
PMC11225860 | Aurkin-A, a TPX2-aurora a small molecule inhibitor disrupts Alisertib-induced polyploidy in aggressive diffuse large B cell lymphoma | Polyploidy is a specific form of aneuploidy caused by whole genome doubling (WGD), which involves successive rounds of replication without mitosis or cytokinesis, resulting in polyploid nuclei or the formation of several nuclei within the same cell . |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.