Abstract
Motivation
Historically, gene expression has been shown to be the most informative data for drug response prediction. Recent evidence suggests that integrating additional omics can improve ...the prediction accuracy which raises the question of how to integrate the additional omics. Regardless of the integration strategy, clinical utility and translatability are crucial. Thus, we reasoned a multi-omics approach combined with clinical datasets would improve drug response prediction and clinical relevance.
Results
We propose MOLI, a multi-omics late integration method based on deep neural networks. MOLI takes somatic mutation, copy number aberration and gene expression data as input, and integrates them for drug response prediction. MOLI uses type-specific encoding sub-networks to learn features for each omics type, concatenates them into one representation and optimizes this representation via a combined cost function consisting of a triplet loss and a binary cross-entropy loss. The former makes the representations of responder samples more similar to each other and different from the non-responders, and the latter makes this representation predictive of the response values. We validate MOLI on in vitro and in vivo datasets for five chemotherapy agents and two targeted therapeutics. Compared to state-of-the-art single-omics and early integration multi-omics methods, MOLI achieves higher prediction accuracy in external validations. Moreover, a significant improvement in MOLI’s performance is observed for targeted drugs when training on a pan-drug input, i.e. using all the drugs with the same target compared to training only on drug-specific inputs. MOLI’s high predictive power suggests it may have utility in precision oncology.
Availability and implementation
https://github.com/hosseinshn/MOLI.
Supplementary information
Supplementary data are available at Bioinformatics online.
More potent targeting of the androgen receptor (AR) in advanced prostate cancer is driving an increased incidence of neuroendocrine prostate cancer (NEPC), an aggressive and treatment-resistant ...AR-negative variant. Its molecular pathogenesis remains poorly understood but appears to require TP53 and RB1 aberration. We modeled the development of NEPC from conventional prostatic adenocarcinoma using a patient-derived xenograft and found that the placental gene PEG10 is de-repressed during the adaptive response to AR interference and subsequently highly upregulated in clinical NEPC. We found that the AR and the E2F/RB pathway dynamically regulate distinct post-transcriptional and post-translational isoforms of PEG10 at distinct stages of NEPC development. In vitro, PEG10 promoted cell-cycle progression from G0/G1 in the context of TP53 loss and regulated Snail expression via TGF-β signaling to promote invasion. Taken together, these findings show the mechanistic relevance of RB1 and TP53 loss in NEPC and suggest PEG10 as a NEPC-specific target.
Display omitted
•Placental gene PEG10 is highly expressed in neuroendocrine prostate cancer (NEPC)•PEG10 is dynamically regulated by AR and E2F/RB during NEPC development•Distinct isoforms of PEG10 promote proliferation and invasion of NEPC cells•PEG10 represents a specific therapeutic target for NEPC
Akamatsu et al. describe involvement of the placental gene PEG10 in driving the proliferative and invasive phenotype of lethal neuroendocrine prostate cancer (NEPC), suggesting PEG10 as a therapeutic target.
Although novel agents targeting the androgen-androgen receptor (AR) axis have altered the treatment paradigm of metastatic castration-resistant prostate cancer (mCRPC), development of therapeutic ...resistance is inevitable. In this study, we examined whether AR gene aberrations detectable in circulating cell-free DNA (cfDNA) are associated with resistance to abiraterone acetate and enzalutamide in mCRPC patients.
Plasma was collected from 62 mCRPC patients ceasing abiraterone acetate (n = 29), enzalutamide (n = 19), or other agents (n = 14) due to disease progression. DNA was extracted and subjected to array comparative genomic hybridization (aCGH) for chromosome copy number analysis, and Roche 454 targeted next-generation sequencing of exon 8 in the AR.
On aCGH, AR amplification was significantly more common in patients progressing on enzalutamide than on abiraterone or other agents (53% vs. 17% vs. 21%, P = 0.02, χ(2)). Missense AR exon 8 mutations were detected in 11 of 62 patients (18%), including the first reported case of an F876L mutation in an enzalutamide-resistant patient and H874Y and T877A mutations in 7 abiraterone-resistant patients. In patients switched onto enzalutamide after cfDNA collection (n = 39), an AR gene aberration (copy number increase and/or an exon 8 mutation) in pretreatment cfDNA was associated with adverse outcomes, including lower rates of PSA decline ≥ 30% (P = 0.013, χ(2)) and shorter time to radiographic/clinical progression (P = 0.010, Cox proportional hazards regression).
AR gene aberrations in cfDNA are associated with resistance to enzalutamide and abiraterone in mCRPC. Our data illustrate that genomic analysis of cfDNA is a minimally invasive method for interrogating mechanisms of therapeutic resistance in mCRPC.
It is increasingly evident that non–protein-coding regions of the genome can give rise to transcripts that form functional layers of the cancer genome. One of most abundant classes in these regions ...is long noncoding RNAs (lncRNAs). They have gained increasing attention in prostate cancer (PCa) and paved the way for a greater understanding of these cryptic regulators in cancer.
To review current research exploring the functional biology of lncRNAs in PCa over the past three decades.
A systematic review was performed using PubMed to search for reports with terms “long noncoding RNA”, “prostate”, and “cancer” over the past 30 yr (1988–2018).
We comprehensively surveyed the literature collected and summarise experiments leading to the characterisation of lncRNAs in PCa. A historical timeline of lncRNA identification is described, where each lncRNA is categorised mechanistically and within the primary areas of carcinogenesis: tumour risk and initiation, tumour promotion, tumour suppression, and tumour treatment resistance. We describe select lncRNAs that exemplify these areas. We also review whether these lncRNAs have a clinical utility in PCa diagnosis, prognosis, and prediction, and as therapeutic targets.
The biology of lncRNA is multifaceted, demonstrating a complex array of molecular and cellular functions. These studies reveal that lncRNAs are involved in every stage of PCa. Their clinical utility for diagnosis, prognosis, and prediction of PCa is well supported, but further evaluation for their therapeutic candidacy is needed. We provide a detailed resource and view inside the lncRNA landscape for other cancer biologists, oncologists, and clinicians.
In this study, we review current knowledge of the non–protein-coding genome in prostate cancer (PCa). We conclude that many of these regions are functional and a source of accurate biomarkers in PCa. With a strong research foundation, they hold promise as future therapeutic targets, yet clinical trials are necessary to determine their intrinsic value to PCa disease management.
This review comprehensively summarises three decades of knowledge on long noncoding RNA biology and research in prostate cancer. It systematically summarises the significant contribution of these master regulators in tumorigenesis and how this translates clinically.
DNA sequencing has identified recurrent mutations that drive the aggressiveness of prostate cancers. Surprisingly, the influence of genomic, epigenomic, and transcriptomic dysregulation on the tumor ...proteome remains poorly understood. We profiled the genomes, epigenomes, transcriptomes, and proteomes of 76 localized, intermediate-risk prostate cancers. We discovered that the genomic subtypes of prostate cancer converge on five proteomic subtypes, with distinct clinical trajectories. ETS fusions, the most common alteration in prostate tumors, affect different genes and pathways in the proteome and transcriptome. Globally, mRNA abundance changes explain only ∼10% of protein abundance variability. As a result, prognostic biomarkers combining genomic or epigenomic features with proteomic ones significantly outperform biomarkers comprised of a single data type.
Display omitted
•A comprehensive proteomic analyses of localized prostate cancers•Integration of all levels of the central dogma (DNA → RNA → protein)•ETS fusions have divergent effects on transcriptome and proteome•Combining genomics and proteomics improves biomarker performance
Sinha et al. determine the proteogenomic landscape of localized, intermediate-risk prostate cancers and show that the presence of ETS gene fusions has one of the strongest effects on the proteome. Prognostic biomarkers that integrate multi-omics significantly outperform those comprised of a single data type.
Abstract
Motivation
The goal of pharmacogenomics is to predict drug response in patients using their single- or multi-omics data. A major challenge is that clinical data (i.e. patients) with drug ...response outcome is very limited, creating a need for transfer learning to bridge the gap between large pre-clinical pharmacogenomics datasets (e.g. cancer cell lines), as a source domain, and clinical datasets as a target domain. Two major discrepancies exist between pre-clinical and clinical datasets: (i) in the input space, the gene expression data due to difference in the basic biology, and (ii) in the output space, the different measures of the drug response. Therefore, training a computational model on cell lines and testing it on patients violates the i.i.d assumption that train and test data are from the same distribution.
Results
We propose Adversarial Inductive Transfer Learning (AITL), a deep neural network method for addressing discrepancies in input and output space between the pre-clinical and clinical datasets. AITL takes gene expression of patients and cell lines as the input, employs adversarial domain adaptation and multi-task learning to address these discrepancies, and predicts the drug response as the output. To the best of our knowledge, AITL is the first adversarial inductive transfer learning method to address both input and output discrepancies. Experimental results indicate that AITL outperforms state-of-the-art pharmacogenomics and transfer learning baselines and may guide precision oncology more accurately.
Availability and implementation
https://github.com/hosseinshn/AITL.
Supplementary information
Supplementary data are available at Bioinformatics online.
Prostate cancer (PCa) is the most common malignant neoplasm among men in many countries. Since most precancerous and cancerous tissues show signs of inflammation, chronic bacterial prostatitis has ...been hypothesized to be a possible etiology. However, establishing a causal relationship between microbial inflammation and PCa requires a comprehensive analysis of the prostate microbiome. The aim of this study was to characterize the microbiome in prostate tissue of PCa patients and investigate its association with tumour clinical characteristics as well as host expression profiles.
The metagenome and metatranscriptome of tumour and the adjacent benign tissues were assessed in 65 Chinese radical prostatectomy specimens. Escherichia, Propionibacterium, Acinetobacter and Pseudomonas were abundant in both metagenome and metatranscriptome, thus constituting the core of the prostate microbiome. The biodiversity of the microbiomes could not be differentiated between the matched tumour/benign specimens or between the tumour specimens of low and high Gleason Scores. The expression profile of ten Pseudomonas genes was strongly correlated with that of eight host small RNA genes; three of the RNA genes may negatively associate with metastasis. Few viruses could be identified from the prostate microbiomes.
This is the first study of the human prostate microbiome employing an integrated metagenomics and metatranscriptomics approach. In this Chinese cohort, both metagenome and metatranscriptome analyses showed a non-sterile microenvironment in the prostate of PCa patients, but we did not find links between the microbiome and local progression of PCa. However, the correlated expression of Pseudomonas genes and human small RNA genes may provide tantalizing preliminary evidence that Pseudomonas infection may impede metastasis.
Targeted agents and immunotherapies promise to transform the treatment of metastatic bladder cancer, but therapy selection will depend on practical tumor molecular stratification. Circulating tumor ...DNA (ctDNA) is established in several solid malignancies as a minimally invasive tool to profile the tumor genome in real-time, but is critically underexplored in bladder cancer.
We applied a combination of whole-exome sequencing and targeted sequencing across 50 bladder cancer driver genes to plasma cell-free DNA (cfDNA) from 51 patients with aggressive bladder cancer, including 37 with metastatic disease.
The majority of patients with metastasis, but only 14% of patients with localized disease, had ctDNA proportions above 2% of total cfDNA (median 16.5%, range 3.9%-72.6%). Twelve percent of estimable samples had evidence of genome hypermutation. We reveal an aggressive mutational landscape in metastatic bladder cancer with 95% of patients harboring deleterious alterations to
, or
, and 70% harboring a mutation or disrupting rearrangement affecting chromatin modifiers such as
Targetable alterations in MAPK/ERK or PI3K/AKT/mTOR pathways were robustly detected, including amplification of
(20% of patients) and activating hotspot mutations in
A (20%), with the latter mutually exclusive to truncating mutations in
A novel
gene fusion was identified in consecutive samples from one patient.
Our study demonstrates that ctDNA provides a practical and cost-effective snapshot of driver gene status in metastatic bladder cancer. The identification of a wide spectrum of clinically informative somatic alterations nominates ctDNA as a tool to dissect disease pathogenesis and guide therapy selection in patients with metastatic bladder cancer.
.