Archive for the ‘Computational Biology/Systems and Bioinformatics’ Category

Deep Learning extracts Histopathological Patterns and accurately discriminates 28 Cancer and 14 Normal Tissue Types: Pan-cancer Computational Histopathology Analysis

Posted in Artificial Intelligence Applications in Health Care, Artificial Intelligence in CANCER, Artificial Intelligence in Health Care - Tools & Innovations, Artificial Intelligence in Medicine - Application for Diagnosis, CANCER BIOLOGY & Innovations in Cancer Therapy, Computational Biology/Systems and Bioinformatics, Computational Histopathology, Deep Learning in Pathology, Genome Biology, Meta-analysis of transcriptome data on October 28, 2019| Leave a Comment »

Deep Learning extracts Histopathological Patterns and accurately discriminates 28 Cancer and 14 Normal Tissue Types: Pan-cancer Computational Histopathology Analysis

Reporter: Aviva Lev-Ari, PhD, RN

Pan-cancer computational histopathology reveals mutations, tumor composition and prognosis

Yu Fu1, Alexander W Jung1, Ramon Viñas Torne1, Santiago Gonzalez1,2, Harald Vöhringer1, Mercedes Jimenez-Linan3, Luiza Moore3,4, and Moritz Gerstung#1,5 # to whom correspondence should be addressed 1) European Molecular Biology Laboratory, European Bioinformatics Institute (EMBL-EBI), Hinxton, UK. 2) Current affiliation: Institute for Research in Biomedicine (IRB Barcelona), Parc Científic de Barcelona, Barcelona, Spain. 3) Department of Pathology, Addenbrooke’s Hospital, Cambridge, UK. 4) Wellcome Sanger Institute, Hinxton, UK 5) European Molecular Biology Laboratory, Genome Biology Unit, Heidelberg, Germany.

Correspondence:

Dr Moritz Gerstung European Molecular Biology Laboratory European Bioinformatics Institute (EMBL-EBI) Hinxton, CB10 1SA UK. Tel: +44 (0) 1223 494636 E-mail: moritz.gerstung@ebi.ac.uk

Abstract

Pan-cancer computational histopathology reveals mutations, tumor composition and prognosis

Here we use deep transfer learning to quantify histopathological patterns across 17,396 H&E stained histopathology image slides from 28 cancer types and correlate these with underlying genomic and transcriptomic data. Pan-cancer computational histopathology (PC-CHiP) classifies the tissue origin across organ sites and provides highly accurate, spatially resolved tumor and normal distinction within a given slide. The learned computational histopathological features correlate with a large range of recurrent genetic aberrations, including whole genome duplications (WGDs), arm-level copy number gains and losses, focal amplifications and deletions as well as driver gene mutations within a range of cancer types. WGDs can be predicted in 25/27 cancer types (mean AUC=0.79) including those that were not part of model training. Similarly, we observe associations with 25% of mRNA transcript levels, which enables to learn and localise histopathological patterns of molecularly defined cell types on each slide. Lastly, we find that computational histopathology provides prognostic information augmenting histopathological subtyping and grading in the majority of cancers assessed, which pinpoints prognostically relevant areas such as necrosis or infiltrating lymphocytes on each tumour section. Taken together, these findings highlight the large potential of PC-CHiP to discover new molecular and prognostic associations, which can augment diagnostic workflows and lay out a rationale for integrating molecular and histopathological data.

SOURCE

https://www.biorxiv.org/content/10.1101/813543v1

Key points

● Pan-cancer computational histopathology analysis with deep learning extracts histopathological patterns and accurately discriminates 28 cancer and 14 normal tissue types

● Computational histopathology predicts whole genome duplications, focal amplifications and deletions, as well as driver gene mutations

● Wide-spread correlations with gene expression indicative of immune infiltration and proliferation

● Prognostic information augments conventional grading and histopathology subtyping in the majority of cancers

Discussion

Here we presented PC-CHiP, a pan-cancer transfer learning approach to extract computational histopathological features across 42 cancer and normal tissue types and their genomic, molecular and prognostic associations. Histopathological features, originally derived to classify different tissues, contained rich histologic and morphological signals predictive of a range of genomic and transcriptomic changes as well as survival. This shows that computer vision not only has the capacity to highly accurately reproduce predefined tissue labels, but also that this quantifies diverse histological patterns, which are predictive of a broad range of genomic and molecular traits, which were not part of the original training task. As the predictions are exclusively based on standard H&E-stained tissue sections, our analysis highlights the high potential of computational histopathology to digitally augment existing histopathological workflows. The strongest genomic associations were found for whole genome duplications, which can in part be explained by nuclear enlargement and increased nuclear intensities, but seemingly also stems from tumour grade and other histomorphological patterns contained in the high-dimensional computational histopathological features. Further, we observed associations with a range of chromosomal gains and losses, focal deletions and amplifications as well as driver gene mutations across a number of cancer types. These data demonstrate that genomic alterations change the morphology of cancer cells, as in the case of WGD, but possibly also that certain aberrations preferentially occur in distinct cell types, reflected by the tumor histology. Whatever is the cause or consequence in this equation, these associations lay out a route towards genomically defined histopathology subtypes, which will enhance and refine conventional assessment. Further, a broad range of transcriptomic correlations was observed reflecting both immune cell infiltration and cell proliferation that leads to higher tumor densities. These examples illustrated the remarkable property that machine learning does not only establish novel molecular associations from pre-computed histopathological feature sets but also allows the localisation of these traits within a larger image. While this exemplifies the power of a large scale data analysis to detect and localise recurrent patterns, it is probably not superior to spatially annotated training data. Yet such data can, by definition, only be generated for associations which are known beforehand. This appears straightforward, albeit laborious, for existing histopathology classifications, but more challenging for molecular readouts. Yet novel spatial transcriptomic44,45 and sequencing technologies46 bring within reach spatially matched molecular and histopathological data, which would serve as a gold standard in combining imaging and molecular patterns. Across cancer types, computational histopathological features showed a good level of prognostic relevance, substantially improving prognostic accuracy over conventional grading and histopathological subtyping in the majority of cancers. It is this very remarkable that such predictive It is made available under a CC-BY-NC 4.0 International license. (which was not peer-reviewed) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. bioRxiv preprint first posted online Oct. 25, 2019; doi: http://dx.doi.org/10.1101/813543. The copyright holder for this preprint signals can be learned in a fully automated fashion. Still, at least at the current resolution, the improvement over a full molecular and clinical workup was relatively small. This might be a consequence of the far-ranging relations between histopathology and molecular phenotypes described here, implying that histopathology is a reflection of the underlying molecular alterations rather than an independent trait. Yet it probably also highlights the challenges of unambiguously quantifying histopathological signals in – and combining signals from – individual areas, which requires very large training datasets for each tumour entity. From a methodological point of view, the prediction of molecular traits can clearly be improved. In this analysis, we adopted – for the reason of simplicity and to avoid overfitting – a transfer learning approach in which an existing deep convolutional neural network, developed for classification of everyday objects, was fine tuned to predict cancer and normal tissue types. The implicit imaging feature representation was then used to predict molecular traits and outcomes. Instead of employing this two-step procedure, which risks missing patterns irrelevant for the initial classification task, one might directly employ either training on the molecular trait of interest, or ideally multi-objective learning. Further improvement may also be related to the choice of the CNN architecture. Everyday images have no defined scale due to a variable z-dimension; therefore, the algorithms need to be able to detect the same object at different sizes. This clearly is not the case for histopathology slides, in which one pixel corresponds to a defined physical size at a given magnification. Therefore, possibly less complex CNN architectures may be sufficient for quantitative histopathology analyses, and also show better generalisation. Here, in our proof-of-concept analysis, we observed a considerable dependence of the feature representation on known and possibly unknown properties of our training data, including the image compression algorithm and its parameters. Some of these issues could be overcome by amending and retraining the network to isolate the effect of confounding factors and additional data augmentation. Still, given the flexibility of deep learning algorithms and the associated risk of overfitting, one should generally be cautious about the generalisation properties and critically assess whether a new image is appropriately represented. Looking forward, our analyses revealed the enormous potential of using computer vision alongside molecular profiling. While the eye of a trained human may still constitute the gold standard for recognising clinically relevant histopathological patterns, computers have the capacity to augment this process by sifting through millions of images to retrieve similar patterns and establish associations with known and novel traits. As our analysis showed this helps to detect histopathology patterns associated with a range of genomic alterations, transcriptional signatures and prognosis – and highlight areas indicative of these traits on each given slide. It is therefore not too difficult to foresee how this may be utilised in a computationally augmented histopathology workflow enabling more precise and faster diagnosis and prognosis. Further, the ability to quantify a rich set of histopathology patterns lays out a path to define integrated histopathology and molecular cancer subtypes, as recently demonstrated for colorectal cancers47 .

Lastly, our analyses provide It is made available under a CC-BY-NC 4.0 International license. (which was not peer-reviewed) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity.

bioRxiv preprint first posted online Oct. 25, 2019; doi: http://dx.doi.org/10.1101/813543.

The copyright holder for this preprint proof-of-concept for these principles and we expect them to be greatly refined in the future based on larger training corpora and further algorithmic refinements.

SOURCE

https://www.biorxiv.org/content/biorxiv/early/2019/10/25/813543.full.pdf

Single-cell RNA-seq helps in finding intra-tumoral heterogeneity in pancreatic cancer

Posted in Artificial Intelligence in CANCER, BioIT: BioInformatics, NGS, Clinical & Translational, Pharmaceutical R&D Informatics, Clinical Genomics, Cancer Informatics, Biological Networks, Biological Networks, Gene Regulation and Evolution, Cancer - General, Cancer and Current Therapeutics, CANCER BIOLOGY & Innovations in Cancer Therapy, Cancer Genomics, Cancer Informatics, Cancer Prevention: Research & Programs, Cancer Screening, Cancer-Immune Interactions, Cell Biology, Cell Biology, Signaling & Cell Circuits, Immuno-Oncology & Genomics, Meta-analysis of transcriptome data, mRNA platform in Drug DIscovery, Pancreatic adenocarcinoma classifier, Pancreatic cancer, Precision Cancer Medicine, RNA Biology, RNA Biology, Cancer and Therapeutics, Single-cell sequencing, Transcriptomics, tagged heterogeneity, Pancreatic cancer, pancreatic ductal adenocarcinoma, single-cell RNA sequencing, transcription on July 21, 2019| Leave a Comment »

Single-cell RNA-seq helps in finding intra-tumoral heterogeneity in pancreatic cancer

Reporter and Curator: Dr. Sudipta Saha, Ph.D.

Pancreatic cancer is a significant cause of cancer mortality; therefore, the development of early diagnostic strategies and effective treatment is essential. Improvements in imaging technology, as well as use of biomarkers are changing the way that pancreas cancer is diagnosed and staged. Although progress in treatment for pancreas cancer has been incremental, development of combination therapies involving both chemotherapeutic and biologic agents is ongoing.

Cancer is an evolutionary disease, containing the hallmarks of an asexually reproducing unicellular organism subject to evolutionary paradigms. Pancreatic ductal adenocarcinoma (PDAC) is a particularly robust example of this phenomenon. Genomic features indicate that pancreatic cancer cells are selected for fitness advantages when encountering the geographic and resource-depleted constraints of the microenvironment. Phenotypic adaptations to these pressures help disseminated cells to survive in secondary sites, a major clinical problem for patients with this disease.

The immune system varies in cell types, states, and locations. The complex networks, interactions, and responses of immune cells produce diverse cellular ecosystems composed of multiple cell types, accompanied by genetic diversity in antigen receptors. Within this ecosystem, innate and adaptive immune cells maintain and protect tissue function, integrity, and homeostasis upon changes in functional demands and diverse insults. Characterizing this inherent complexity requires studies at single-cell resolution. Recent advances such as massively parallel single-cell RNA sequencing and sophisticated computational methods are catalyzing a revolution in our understanding of immunology.

PDAC is the most common type of pancreatic cancer featured with high intra-tumoral heterogeneity and poor prognosis. In the present study to comprehensively delineate the PDAC intra-tumoral heterogeneity and the underlying mechanism for PDAC progression, single-cell RNA-seq (scRNA-seq) was employed to acquire the transcriptomic atlas of 57,530 individual pancreatic cells from primary PDAC tumors and control pancreases. The diverse malignant and stromal cell types, including two ductal subtypes with abnormal and malignant gene expression profiles respectively, were identified in PDAC.

The researchers found that the heterogenous malignant subtype was composed of several subpopulations with differential proliferative and migratory potentials. Cell trajectory analysis revealed that components of multiple tumor-related pathways and transcription factors (TFs) were differentially expressed along PDAC progression. Furthermore, it was found a subset of ductal cells with unique proliferative features were associated with an inactivation state in tumor-infiltrating T cells, providing novel markers for the prediction of antitumor immune response. Together, the findings provided a valuable resource for deciphering the intra-tumoral heterogeneity in PDAC and uncover a connection between tumor intrinsic transcriptional state and T cell activation, suggesting potential biomarkers for anticancer treatment such as targeted therapy and immunotherapy.

References:

https://www.ncbi.nlm.nih.gov/pubmed/31273297

https://www.ncbi.nlm.nih.gov/pubmed/21491194

https://www.ncbi.nlm.nih.gov/pubmed/27444064

https://www.ncbi.nlm.nih.gov/pubmed/28983043

https://www.ncbi.nlm.nih.gov/pubmed/24976721

https://www.ncbi.nlm.nih.gov/pubmed/27693023

Read Full Post »

scPopCorn: A New Computational Method for Subpopulation Detection and their Comparative Analysis Across Single-Cell Experiments

Posted in Advanced Computing Platform, Artificial Intelligence Applications in Health Care, Artificial Intelligence in Health Care - Tools & Innovations, Artificial Intelligence in Medicine - Applications in Therapeutics, Big Data & Analytics, BioIT: BioInformatics, BioIT: BioInformatics, NGS, Clinical & Translational, Pharmaceutical R&D Informatics, Clinical Genomics, Cancer Informatics, Biotechnology, Cancer Genomics, Cardiovascular Pharmacogenomics, Cell Biology, Clinical Genomics, Computational Biology/Systems and Bioinformatics, Data Science, Genome Biology, Genomic Analysis of EKG Electrophysiology Genomics, Genomic Endocrinology, Genomic Expression, Genomic Testing: Methodology for Diagnosis, Genomics Pharmacy, Immuno-Oncology & Genomics, Nutrigenomics, Personalized and Precision Medicine & Genomic Research, Pharmacogenomics, Preimplantation Genetic Diagnosis and Reproductive Genomics, Reproductive Andrology, Embryology, Genomic Endocrinology, Preimplantation Genetic Diagnosis and Reproductive Genomics, RNA Biology, RNA Biology, Cancer and Therapeutics, Single Cell Genomics, Single-cell sequencing, Uncategorized, tagged comparative analysis, Computational Biology/Systems and Bioinformatics, computational genomics, scPopCorn, single-cell RNA sequencing, subpopulation on July 16, 2019| Leave a Comment »

scPopCorn: A New Computational Method for Subpopulation Detection and their Comparative Analysis Across Single-Cell Experiments

Reporter and Curator: Dr. Sudipta Saha, Ph.D.

Present day technological advances have facilitated unprecedented opportunities for studying biological systems at single-cell level resolution. For example, single-cell RNA sequencing (scRNA-seq) enables the measurement of transcriptomic information of thousands of individual cells in one experiment. Analyses of such data provide information that was not accessible using bulk sequencing, which can only assess average properties of cell populations. Single-cell measurements, however, can capture the heterogeneity of a population of cells. In particular, single-cell studies allow for the identification of novel cell types, states, and dynamics.

One of the most prominent uses of the scRNA-seq technology is the identification of subpopulations of cells present in a sample and comparing such subpopulations across samples. Such information is crucial for understanding the heterogeneity of cells in a sample and for comparative analysis of samples from different conditions, tissues, and species. A frequently used approach is to cluster every dataset separately, inspect marker genes for each cluster, and compare these clusters in an attempt to determine which cell types were shared between samples. This approach, however, relies on the existence of predefined or clearly identifiable marker genes and their consistent measurement across subpopulations.

Although the aligned data can then be clustered to reveal subpopulations and their correspondence, solving the subpopulation-mapping problem by performing global alignment first and clustering second overlooks the original information about subpopulations existing in each experiment. In contrast, an approach addressing this problem directly might represent a more suitable solution. So, keeping this in mind the researchers developed a computational method, single-cell subpopulations comparison (scPopCorn), that allows for comparative analysis of two or more single-cell populations.

The performance of scPopCorn was tested in three distinct settings. First, its potential was demonstrated in identifying and aligning subpopulations from single-cell data from human and mouse pancreatic single-cell data. Next, scPopCorn was applied to the task of aligning biological replicates of mouse kidney single-cell data. scPopCorn achieved the best performance over the previously published tools. Finally, it was applied to compare populations of cells from cancer and healthy brain tissues, revealing the relation of neoplastic cells to neural cells and astrocytes. Consequently, as a result of this integrative approach, scPopCorn provides a powerful tool for comparative analysis of single-cell populations.

This scPopCorn is basically a computational method for the identification of subpopulations of cells present within individual single-cell experiments and mapping of these subpopulations across these experiments. Different from other approaches, scPopCorn performs the tasks of population identification and mapping simultaneously by optimizing a function that combines both objectives. When applied to complex biological data, scPopCorn outperforms previous methods. However, it should be kept in mind that scPopCorn assumes the input single-cell data to consist of separable subpopulations and it is not designed to perform a comparative analysis of single cell trajectories datasets that do not fulfill this constraint.

Several innovations developed in this work contributed to the performance of scPopCorn. First, unifying the above-mentioned tasks into a single problem statement allowed for integrating the signal from different experiments while identifying subpopulations within each experiment. Such an incorporation aids the reduction of biological and experimental noise. The researchers believe that the ideas introduced in scPopCorn not only enabled the design of a highly accurate identification of subpopulations and mapping approach, but can also provide a stepping stone for other tools to interrogate the relationships between single cell experiments.

References:

https://www.sciencedirect.com/science/article/pii/S2405471219301887

https://www.tandfonline.com/doi/abs/10.1080/23307706.2017.1397554

https://ieeexplore.ieee.org/abstract/document/4031383

https://genomebiology.biomedcentral.com/articles/10.1186/s13059-016-0927-y

https://www.sciencedirect.com/science/article/pii/S2405471216302666

Read Full Post »

Real Time @BIOConvention #BIO2019:#Bitcoin Your Data! From Trusted Pharma Silos to Trustless Community-Owned Blockchain-Based Precision Medicine Data Trials

Posted in and Bioethics, Computational Biology/Systems and Bioinformatics, Conference Coverage with Social Media, Diagnostics and Lab Tests, Health Law & Patient Safety, HealthCare IT, Intellectual Property, Innovations, Commercialization, Investment in technological breakthrough, tagged #mhealth, bitcoin, blockchain, Clinical Trials, healthIT on June 5, 2019| Leave a Comment »

Real Time @BIOConvention #BIO2019:#Bitcoin Your Data! From Trusted Pharma Silos to Trustless Community-Owned Blockchain-Based Precision Medicine Data Trials

Reporter: Stephen J Williams, PhD @StephenJWillia2

Speakers

Sanjeev Wadhwa Chief Executive Officer at Life Singularity, Inc.

James CanterburyPrincipal at EY

John KamaraFounder at Alpha Health

Dr. David KoepsellCEO at Encrypgen

Katherine KuzmeskasCEO at SimplyVitalHealth.com

Kamal ObbadCEO at Nebula Genomics

As care for lifestyle-driven chronic diseases expands in scope, prevention and recovery are becoming the new areas of focus. Building a precision medicine foundation that will promote ownership of individuals’ health data and allow for sharing and trading of this data could prove a great blockchain.

At its core, blockchain may offer the potential of a shared platform that decentralizes healthcare interactions ensuring access control, authenticity and integrity, while presenting the industry with radical possibilities for value-based care and reimbursement models. Panelists will explore these new discoveries as well as look to answer lingering questions, such as: are we off to a “trustless” information model underpinned by Bitcoin cryptocurrency, where no central authority validates the transactions in the ledger, and anyone whose computers can do the required math can join to mine and add blocks to your data? Would smart contracts begin to incentivize “rational” behaviors where consumers respond in a manner that makes their data interesting?

Moderator: Cybersecurity is extremely important in the minds of healthcare CEOs. CEO of Kaiser Permenente has listed this as one of main concerns for his company.

Sanjeey of Singularity: There are Very few companies in this space. Singularity have collected thousands of patient data. They wanted to do predictive health care, where a patient will know beforehand what health problems and issues to expect. Created a program called Virtual Assistant. As data is dynamic, the goal was to provide Virtual Assistant to everyone.

Benefits of blockchain: secure, simple to update, decentralized data; patient can control their own data, who sees it and monetize it.

Nebular Genetics: Company was founded by Dr. George Church, who had pioneered the next generation sequencing (NGS) methodology. The company goal is to make genomics available to all but this currently is not the case as NGS is not being used as frequently.

The problem is a data problem:

data not organized
data too parsed
data not accessible

Blockchain may be able to alleviate the accessibiltiy problem. Pharma is very interested in the data but expensive to collect. In addition many companies just do large scale but low depth sequencing. For example 23andme (which had recently made a big deal with Lilly for data) only sequences about 1% of genome.

There are two types of genome sequencing companies

large scale and low depth – like 23andme
smaller scale but higher depth – like DECODE and some of the EU EXOME sequencing efforts like the 1000 Project

Simply Vital Health: Harnesses blockchain to combat ineffeciencies in hospital records. They tackle the costs after acute care so increase the value based care. Most of healthcare is concentrated on the top earners and little is concentrated on the majority less affluent and poor. On addressing HIPAA compliance issues: they decided to work with HIPAA and comply but will wait for this industry to catch up so the industry as a whole can lobby to affect policy change required for blockchain technology to work efficiently in this arena. They will only work with known vendors: VERY Important to know where the data is kept and who are controlling the servers you are using. With other blockchain like Etherium or Bitcoin, the servers are anonymous.

Encrypgen: generates new blockchain for genomic data and NGS companies.

Please follow LIVE on TWITTER using the following @ handles and # hashtags:

@Handles

@pharma_BI

@AVIVA1950

@BIOConvention

# Hashtags

#BIO2019 (official meeting hashtag)

#itstartswithone

#RightMixMatters

#blockchain

#bitcoin

#clinicaltrials

Read Full Post »

Real Time Coverage @BIOConvention #BIO2019: Issues of Risk and Reproduceability in Translational and Academic Collaboration; 2:30-4:00 June 3 Philadelphia PA

Posted in Computational Biology/Systems and Bioinformatics, Conference Coverage with Social Media, Global Partnering & Biotech Investment, Investment in Technological Breakthrough, Life Sciences Breakthrough Prize, Pharmaceutical Discovery, Pharmaceutical Drug Discovery, Pharmaceutical Industry Competitive Intelligence, Pharmaceutical R&D Informatics, Pharmaceutical R&D Investment, tagged academic advancement, Academic publishing, basic research, Big data, collaboration, data validity, Innovation, Pharmaceutical industry, Public–private partnership, research funding, tech transfer on June 3, 2019| Leave a Comment »

Real Time Coverage @BIOConvention #BIO2019: Issues of Risk and Reproduceability in Translational and Academic Collaboration; 2:30-4:00 June 3 Philadelphia PA

Reporter: Stephen J. Williams, PhD @StephenJWillia2

Derisking Academic Science: The Unmet Need

Translating academic research into products and new therapies is a very risky venture as only 1% of academic research has been successfully translated into successful products.

Speakers

Richard SollSenior Advisor, Strategic Initiatives at WuXi AppTec

Jim AudiaExecutive Director atChicago Biomedical Consortium

Collaboration from Chicago area universities like U of Chicago, Northwestern, etc. First phase was enhance collaboration between universities by funding faculty recruitment and basic research. Access to core facilities across universities. Have expanded to give alternatives to company formation.

Curtis KeithCSO atBlavatnik Biomedical Accelerator, Harvard University

Half of the partnerships from Harvard and companies have been able to spin out viable startups.

Nadim ShohdyAssistant Dean, Therapeutics Alliances atNYU Langone Health

Most academic PI are not as savvy to start a biotech so they bring in biotechs and build project teams as well as developing a team of ex pharma and biotech experts. Derisk as running as one asset project. Partner as early as possible. A third of their pipeline have been successfully partnered. Work with investors and patent attorneys.

Dominique VerhelleSr Director, Strategic Academic Alliances, Center for External Innovation atTakeda

Focused on getting PIs to get to startup. Focused on oncology and vaccines and I/O. The result can be liscensing or partnership. Running around 50 to 60 projects. Creating a new company from these US PI partnerships.

Most projects from Harvard have been therapeutics-based. At Harvard they have a network of investors ($50 million). They screen PI proposals based on translateability and what investors are interested in.

In Chicago they solicit multiple projects but are agnostic on area but as they are limited they are focused on projects that will assist in developing a stronger proposal to investor/funding mechanism.

NYU goes around university doing due diligence reaching out to investigators. They shop around their projects to wet their investors, pharma appetite future funding. At Takeda they have five centers around US. They want to have more input so go into the university with their scientists and discuss ideas.

Challenges:

Takeda: Data Validation very important. Second there may be disconnect with the amount of equity the PI wants in the new company as well as management. Third PIs not aware of all steps in drug development.

Harvard: Pharma and biotech have robust research and academic does not have the size or scope of pharma. PIs must be more diligent on e.g. the compounds they get from a screen… they only focus narrowly

NYU: bring in consultants as PIs don’t understand all the management issues. Need to understand development so they bring in the experts to help them. Pharma he feels have to much risk aversion and none of their PIs want 100% equity.

Chicago: they like to publish at early stage so publication freedom is a challenge

It Starts with One Health: A Guide to Translational Science from Research to Investment

115B, Level 100

Speakers

William KareshExecutive Vice President for Health and Policy atEcoHealth Alliance

Marianne de backerVP, Global M&A Operations and Divestitures (Pharmaceuticals Group), Global Infectious Diseases & Vaccines Business Development atJohnson & Johnson Innovation

Randal J. KirkChief Executive Officer atIntrexon

Dr. Hilary MarstonMedical Officer/Policy Advisor, Office of the Director, NIAID atNational Institute of Health

Tina ParkerProgram Officer – Translational Centers of Excellence and Research Coordination Section Office of Biodefense, Research Resources and Translational Research, DMID/NIAID atNIH

Novel Approaches to Improve Reproducibility in Academia-Industry Collaborations

112AB, Level 100

Speakers

Florence BietrixOperations Manager at EATRIS

Dr. Nathan CoussensSenior Research Scientist at National Center for Advancing Translational Sciences, National Institutes of Health

Leonard FreedmanCSO atFrederick National Laboratory for Cancer Research

Marian Hajduch Scientific Director at IMTM

Weida TongDirector Division of Bioinformatics and Biostatistics at NCTR/FDA

Dr. Freedman: Most scientists responding to Nature survey said yes a reproduceability crisis. The reasons: experimental bias, lack of validation techniques, reagents, and protocols etc.

And as he says there is a great ECONOMIC IMPACT of preclinical reproducability issues: to the tune of $56 billion of irreproducable results (paper published in PLOS Biology). If can find the core drivers of this issue they can solve the problem. STANDARDS are constantly used in various industries however academic research are lagging in developing such standards. Just the problem of cell line authentication is costing $4 billion.

Dr. Cousins: There are multiple high throughput screening (HTS) academic centers around the world (150 in US). So where does the industry go for best practices in assays? Eli Lilly had developed a manual for HTS best practices and in 1984 made publicly available (Assay Guidance Manual). To date there have been constant updates to this manual to incorporate new assays. Workshops have been developed to train scientists in these best practices.

NIH has been developing new programs to address these reproducability issues. Developed a method called

“Ring Testing Initiative” where multiple centers involved in sharing reagents as well as assays and allowing scientists to test at multiple facilities.

Dr.Tong: Reproduceability of Microarrays: As microarrays were the only methodology to do high through put genomics in the early 2000s, and although much research had been performed to standardize and achieve best reproduceability of the microarray technology (determining best practices in spotting RNA on glass slides, hybridization protocols, image analysis) little had been done on evaluating the reproducibility of results obtained from microarray experiments involving biological samples. The advent of Artificial Intelligence and Machine Learning though can be used to help validate microarray results. This was done in a Nature Biotechnology paper (Nature Biotechnology volume28, 827–838 (2010)) by an international consortium, the International MAQC (Microarray Quality Control) Society and can be found here

The MicroArray Quality Control (MAQC)-II study of common practices for the development and validation of microarray-based predictive models

However Dr. Tong feels there is much confusion in how we define reproduceability. Dr. Tong identified a few key points of data reproduceability:

Traceability: what are the practices and procedures from going from point A to point B (steps in a protocol or experimental design)
Repeatability: ability to repeat results within the same laboratory
Replicatablilty: ability to repeat results cross laboratory
Transferability: are the results validated across multiple platforms?

The panel then discussed the role of journals and funders to drive reproduceability in research. They felt that editors have been doing as much as they can do as they receive an end product (the paper) but all agreed funders need to do more to promote data validity, especially in requiring that systematic evaluation and validation of each step in protocols are performed.. There could be more training of PIs with respect to protocol and data validation.

BioInformatic Resources at the Environmental Protection Agency: Tools and Webinars on Toxicity Prediction

Posted in Artificial Intelligence - General, Big Data, BioIT: BioInformatics, Computational Biology/Systems and Bioinformatics, Scientific & Biotech Conferences: Press Coverage, tagged Big data, bioinformatic tools, Bioinformatics, curated database, environmental protection agency, EPA, toxicology on May 11, 2019| Leave a Comment »

BioInformatic Resources at the Environmental Protection Agency: Tools and Webinars on Toxicity Prediction

Curator Stephen J. Williams Ph.D.

New GenRA Module in EPA’s CompTox Dashboard Will Help Predict Potential Chemical Toxicity

Published September 25, 2018

As part of its ongoing computational toxicology research, EPA is developing faster and improved approaches to evaluate chemicals for potential health effects. One commonly applied approach is known as chemical read-across. Read-across uses information about how a chemical with known data behaves to make a prediction about the behavior of another chemical that is “similar” but does not have as much data. Current read-across, while cost-effective, relies on a subjective assessment, which leads to varying predictions and justifications depending on who undertakes and evaluates the assessment.

To reduce uncertainties and develop a more objective approach, EPA researchers have developed an automated read-across tool called Generalized Read-Across (GenRA), and added it to the newest version of the EPA Computational Toxicology Dashboard. The goal of GenRA is to encode as many expert considerations used within current read-across approaches as possible and combine these with data-driven approaches to transition read-across towards a more systematic and data-based method of making predictions.

EPA chemist Dr. Grace Patlewicz says it was this uncertainty that motivated the development of GenRA. “You don’t actually know if you’ve been successful at using read-across to help predict chemical toxicity because it’s a judgement call based on one person versus the next. That subjectivity is something we were trying to move away from.” Patlewicz says.

Since toxicologists and risk assessors are already familiar with read-across, EPA researchers saw value in creating a tool that that was aligned with the current read-across workflow but which addressed uncertainty using data analysis methods in what they call a “harmonized-hybrid workflow.”

In its current form, GenRA lets users find analogues, or chemicals that are similar to their target chemical, based on chemical structural similarity. The user can then select which analogues they want to carry forward into the GenRA prediction by exploring the consistency and concordance of the underlying experimental data for those analogues. Next, the tool predicts toxicity effects of specific repeated dose studies. Then, a plot with these outcomes is generated based on a similarity-weighted activity of the analogue chemicals the user selected. Finally, the user is presented with a data matrix view showing whether a chemical is predicted to be toxic (yes or no) for a chosen set of toxicity endpoints, with a quantitative measure of uncertainty.

The team is also comparing chemicals based on other similarity contexts, such as physicochemical characteristics or metabolic similarity, as well as extending the approach to make quantitative predictions of toxicity.

Patlewicz thinks incorporating other contexts and similarity measures will refine GenRA to make better toxicity predictions, fulfilling the goal of creating a read-across method capable of assessing thousands of chemicals that currently lack toxicity data.

“That’s the direction that we’re going in,” Patlewicz says. “Recognizing where we are and trying to move towards something a little bit more objective, showing how aspects of the current read-across workflow could be refined.”

Learn more at: https://comptox.epa.gov

A listing of EPA Tools for Air Quality Assessment

Tools

Atmospheric Model Evaluation Tool (AMET)
AMET helps in the evaluation of meteorological and air quality simulations.
Benchmark Dose Software (BMDS)
EPA developed the Benchmark Dose Software (BMDS) as a tool to help estimate dose or exposure of a chemical or chemical mixture associated with a given response level. The methodology is used by EPA risk assessors and is fast becoming the world’s standard for dose-response analysis for risk assessments, including air pollution risk assessments.
BenMAP
BenMAP is a Windows-based computer program that uses a Geographic Information System (GIS)-based to estimate the health impacts and economic benefits occurring when populations experience changes in air quality.
Community-Focused Exposure and Risk Screening Tool (C-FERST)
C-FERST is an online tool developed by EPA in collaboration with stakeholders to provide access to resources that can be used with communities to help identify and learn more about their environmental health issues and explore exposure and risk reduction options.
Community Health Vulnerability Index
EPA scientists developed a Community Health Vulnerability Index that can be used to help identify communities at higher health risk from wildfire smoke. Breathing smoke from a nearby wildfire is a health threat, especially for people with lung or heart disease, diabetes and high blood pressure as well as older adults, and those living in communities with poverty, unemployment and other indicators of social stress. Health officials can use the tool, in combination with air quality models, to focus public health strategies on vulnerable populations living in areas where air quality is impaired, either by wildfire smoke or other sources of pollution. The work was published in Environmental Science & Technology.
Critical Loads Mapper Tool
The Critical Loads Mapper Tool can be used to help protect terrestrial and aquatic ecosystems from atmospheric deposition of nitrogen and sulfur, two pollutants emitted from fossil fuel burning and agricultural emissions. The interactive tool provides easy access to information on deposition levels through time; critical loads, which identify thresholds when pollutants have reached harmful levels; and exceedances of these thresholds.
EnviroAtlas
EnviroAtlas provides interactive tools and resources for exploring the benefits people receive from nature or “ecosystem goods and services”. Ecosystem goods and services are critically important to human health and well-being, but they are often overlooked due to lack of information. Using EnviroAtlas, many types of users can access, view, and analyze diverse information to better understand the potential impacts of various decisions.
EPA Air Sensor Toolbox for Citizen Scientists
EPA’s Air Sensor Toolbox for Citizen Scientists provides information and guidance on new low-cost compact technologies for measuring air quality. Citizens are interested in learning more about local air quality where they live, work and play. EPA’s Toolbox includes information about: Sampling methodologies; Calibration and validation approaches; Measurement methods options; Data interpretation guidelines; Education and outreach; and Low cost sensor performance information.
ExpoFIRST
The Exposure Factors Interactive Resource for Scenarios Tool (ExpoFIRST) brings data from EPA’s Exposure Factors Handbook: 2011 Edition (EFH) to an interactive tool that maximizes flexibility and transparency for exposure assessors. ExpoFIRST represents a significant advance for regional, state, and local scientists in performing and documenting calculations for community and site-specific exposure assessments, including air pollution exposure assessments.
EXPOsure toolbox (ExpoBox)
This is a toolbox created to assist individuals from within government, industry, academia, and the general public with assessing exposure, including exposure to air contaminants, fate and transport processes of air pollutants and their potential exposure concentrations. It is a compendium of exposure assessment tools that links to guidance documents, databases, models, reference materials, and other related resources.
Federal Reference & Federal Equivalency Methods
EPA scientists develop and evaluate Federal Reference Methods and Federal Equivalency Methods for accurately and reliably measuring six primary air pollutants in outdoor air. These methods are used by states and other organizations to assess implementation actions needed to attain National Ambient Air Quality Standards.
Fertilizer Emission Scenario Tool for CMAQ (FEST-C)
FEST-C facilitates the definition and simulation of new cropland farm management system scenarios or editing of existing scenarios to drive Environmental Policy Integrated Climate model (EPIC) simulations. For the standard 12km continental Community Multi-Scale Air Quality model (CMAQ) domain, this amounts to about 250,000 simulations for the U.S. alone. It also produces gridded daily EPIC weather input files from existing hourly Meteorology-Chemistry Interface Processor (MCIP) files, transforms EPIC output files to CMAQ-ready input files and links directly to Visual Environment for Rich Data Interpretation (VERDI) for spatial visualization of input and output files. The December 2012 release will perform all these functions for any CMAQ grid scale or domain.
Instruction Guide and Macro Analysis Tool for Community-led Air Monitoring
EPA has developed two tools for evaluating the performance of low-cost sensors and interpreting the data they collect to help citizen scientists, communities, and professionals learn about local air quality.
Integrated Climate and Land use Scenarios (ICLUS)
Climate change and land-use change are global drivers of environmental change. Impact assessments frequently show that interactions between climate and land-use changes can create serious challenges for aquatic ecosystems, water quality, and air quality. Population projections to 2100 were used to model the distribution of new housing across the landscape. In addition, housing density was used to estimate changes in impervious surface cover. A final report, datasets, the ICLUS+ Web Viewer and ArcGIS tools are available.
Indoor Semi-Volatile Organic Compound (i-SVOC)
i-SVOC Version 1.0 is a general-purpose software application for dynamic modeling of the emission, transport, sorption, and distribution of semi-volatile organic compounds (SVOCs) in indoor environments. i-SVOC supports a variety of uses, including exposure assessment and the evaluation of mitigation options. SVOCs are a diverse group of organic chemicals that can be found in: Many are also present in indoor air, where they tend to bind to interior surfaces and particulate matter (dust).
- Pesticides;
- Ingredients in cleaning agents and personal care products;
- Additives to vinyl flooring, furniture, clothing, cookware, food packaging, and electronics.
Municipal Solid Waste Decision Support Tool (MSW DST)EXIT
This tool is designed to aid solid waste planners in evaluating the cost and environmental aspects of integrated municipal solid waste management strategies. The tool is the result of collaboration between EPA and RTI International and its partners.
Optical Noise-Reduction Averaging (ONA) Program Improves Black Carbon Particle Measurements Using Aethalometers
ONA is a program that reduces noise in real-time black carbon data obtained using Aethalometers. Aethalometers optically measure the concentration of light absorbing or “black” particles that accumulate on a filter as air flows through it. These particles are produced by incomplete fossil fuel, biofuel and biomass combustion. Under polluted conditions, they appear as smoke or haze.
RETIGO tool
Real Time Geospatial Data Viewer (RETIGO) is a free, web-based tool that shows air quality data that are collected while in motion (walking, biking or in a vehicle). The tool helps users overcome technical barriers to exploring air quality data. After collecting measurements, citizen scientists and other users can import their own data and explore the data on a map.
Remote Sensing Information Gateway (RSIG)
RSIG offers a new way for users to get the multi-terabyte, environmental datasets they want via an interactive, Web browser-based application. A file download and parsing process that now takes months will be reduced via RSIG to minutes.
Simulation Tool Kit for Indoor Air Quality and Inhalation Exposure (IAQX)
IAQX version 1.1 is an indoor air quality (IAQ) simulation software package that complements and supplements existing indoor air quality simulation (IAQ) programs. IAQX is for advanced users who have experience with exposure estimation, pollution control, risk assessment, and risk management. There are many sources of indoor air pollution, such as building materials, furnishings, and chemical cleaners. Since most people spend a large portion of their time indoors, it is important to be able to estimate exposure to these pollutants. IAQX helps users analyze the impact of pollutant sources and sinks, ventilation, and air cleaners. It performs conventional IAQ simulations to calculate the pollutant concentration and/or personal exposure as a function of time. It can also estimate adequate ventilation rates based on user-provided air quality criteria. This is a unique feature useful for product stewardship and risk management.
Spatial Allocator
The Spatial Allocator provides tools that could be used by the air quality modeling community to perform commonly needed spatial tasks without requiring the use of a commercial Geographic Information System (GIS).
Traceability Protocol for Assay and Certification of Gaseous Calibration Standards
This is used to certify calibration gases for ambient and continuous emission monitors. It specifies methods for assaying gases and establishing traceability to National Institute of Standards and Technology (NIST) reference standards. Traceability is required under EPA ambient and continuous emission monitoring regulations.
Watershed Deposition Mapping Tool (WDT)
WDT provides an easy to use tool for mapping the deposition estimates from CMAQ to watersheds to provide the linkage of air and water needed for TMDL (Total Maximum Daily Load) and related nonpoint-source watershed analyses.
Visual Environment for Rich Data Interpretation (VERDI)
VERDI is a flexible, modular, Java-based program for visualizing multivariate gridded meteorology, emissions, and air quality modeling data created by environmental modeling systems such as CMAQ and the Weather Research and Forecasting (WRF) model.

Databases

Air Quality Data for the CDC National Environmental Public Health Tracking Network
EPA’s Exposure Research scientists are collaborating with the Centers for Disease Control and Prevention (CDC) on a CDC initiative to build a National Environmental Public Health Tracking (EPHT) network. Working with state, local and federal air pollution and health agencies, the EPHT program is facilitating the collection, integration, analysis, interpretation, and dissemination of data from environmental hazard monitoring, and from human exposure and health effects surveillance. These data provide scientific information to develop surveillance indicators, and to investigate possible relationships between environmental exposures, chronic disease, and other diseases, that can lead to interventions to reduce the burden of theses illnesses. An important part of the initiative is air quality modeling estimates and air quality monitoring data, combined through Bayesian modeling that can be linked with health outcome data.
EPAUS9R – An Energy Systems Database for use with the Market Allocation (MARKAL) Model
The EPAUS9r is a regional database representation of the United States energy system. The database uses the MARKAL model. MARKAL is an energy system optimization model used by local and federal governments, national and international communities and academia. EPAUS9r represents energy supply, technology, and demand throughout the major sectors of the U.S. energy system.
Fused Air Quality Surfaces Using Downscaling
This database provides access to the most recent O3 and PM2.5 surfaces datasets using downscaling.
Health & Environmental Research Online (HERO)
HERO provides access to scientific literature used to support EPA’s integrated science assessments, including the Integrated Science Assessments (ISA) that feed into the National Ambient Air Quality (NAAQS) reviews.
SPECIATE 4.5 Database
SPECIATE is a repository of volatile organic gas and particulate matter (PM) speciation profiles of air pollution sources.

A listing of EPA Tools and Databases for Water Contaminant Exposure Assessment

Exposure and Toxicity

EPA ExpoBox (A Toolbox for Exposure Assessors)
This toolbox assists individuals from within government, industry, academia, and the general public with assessing exposure from multiple media, including water and sediment. It is a compendium of exposure assessment tools that links to guidance documents, databases, models, reference materials, and other related resources.

Chemical and Product Categories (CPCat) Database
CPCat is a database containing information mapping more than 43,000 chemicals to a set of terms categorizing their usage or function. The comprehensive list of chemicals with associated categories of chemical and product use was compiled from publically available sources. Unique use category taxonomies from each source are mapped onto a single common set of approximately 800 terms. Users can search for chemicals by chemical name, Chemical Abstracts Registry Number, or by CPCat terms associated with chemicals.

A listing of EPA Tools and Databases for Chemical Toxicity Prediction & Assessment

Sequence Alignment to Predict Across Species Susceptibility (SeqAPASS)
SeqAPASS is a fast, online screening tool that allows researchers and regulators to extrapolate toxicity information across species. For some species, such as humans, mice, rats, and zebrafish, the EPA has a large amount of data regarding their toxicological susceptibility to various chemicals. However, the toxicity data for numerous other plants and animals is very limited. SeqAPASS extrapolates from these data rich model organisms to thousands of other non-target species to evaluate their specific potential chemical susceptibility.

Aggregated Computational Toxicology Resource (ACToR)
ACToR is an online warehouse of publicly available chemical toxicity data and can be used to find all publicly available data about potential chemical risks to human health and the environment. ACToR aggregates data from over 1000 public sources on over 500,000 environmental chemicals searchable by chemical name, other identifiers and by chemical structure.
Chemical and Product Category Database
Contains information on how chemicals are used in consumer products. It includes information mapping over 43,000 chemicals to a set of terms categorizing their usage in consumer products or function.
Distributed Structure Searchable Toxicity Database (DSSTox)
Publishes downloadable, structure-searchable, standardized chemical structure files associated with chemical inventories or toxicity data sets of environmental relevance.
Downloadable Computational Toxicology Data
All publicly available computational toxicology data.
Ecotoxicology Database (EcoTox)
Source for finding single chemical toxicity data for aquatic life, terrestrial plants and wildlife.
Toxicity Reference Database (ToxRefDB)
ToxRefDB contains thousands of animal toxicity studies on hundreds of chemicals from 30 years worth of animal toxicity studies.
Toxicity Forecaster Data (ToxCast)
ToxCast data includes high-throughput screening data on thousands of chemicals.

A listing of EPA Webinar and Literature on Bioinformatic Tools and Projects

Comparative Bioinformatics Applications for Developmental Toxicology

Discuss how the US EPA/NCCT is trying to solve the problem of too many chemicals, too high cost, and too much biological uncertainty Discuss the solution the ToxCast Program is proposing; a data-rich system to screen, classify and rank chemicals for further evaluation

https://cfpub.epa.gov/si/si_public_record_report.cfm?Lab=NCCT&dirEntryId=186844

CHEMOINFORMATIC AND BIOINFORMATIC CHALLENGES AT THE US ENVIRONMENTAL PROTECTION AGENCY.

This presentation will provide an overview of both the scientific program and the regulatory activities related to computational toxicology. This presentation will provide an overview of both the scientific program and the regulatory activities related to computational toxicology.

https://cfpub.epa.gov/si/si_public_record_report.cfm?Lab=NCCT&dirEntryId=154013

How Can We Use Bioinformatics to Predict Which Agents Will Cause Birth Defects?

The availability of genomic sequences from a growing number of human and model organisms has provided an explosion of data, information, and knowledge regarding biological systems and disease processes. High-throughput technologies such as DNA and protein microarray biochips are now standard tools for probing the cellular state and determining important cellular behaviors at the genomic/proteomic levels. While these newer technologies are beginning to provide important information on cellular reactions to toxicant exposure (toxicogenomics), a major challenge that remains is the formulation of a strategy to integrate transcript, protein, metabolite, and toxicity data. This integration will require new concepts and tools in bioinformatics. The U.S. National Library of Medicine’s Pubmed site includes 19 million citations and abstracts and continues to grow. The BDSM team is now working on assembling the literature’s unstructured data into a structured database and linking it to BDSM within a system that can then be used for testing and generating new hypotheses. This effort will generate data bases of entities (such as genes, proteins, metabolites, gene ontology processes) linked to PubMed identifiers/abstracts and providing information on the relationships between them. The end result will be an online/standalone tool that will help researchers to focus on the papers most relevant to their query and uncover hidden connections and obvious information gaps.

https://cfpub.epa.gov/si/si_public_record_report.cfm?Lab=NCCT&dirEntryId=227345

ADVANCED PROTEOMICS AND BIOINFORMATICS TOOLS IN TOXICOLOGY RESEARCH: OVERCOMING CHALLENGES TO PROVIDE SIGNIFICANT RESULTS

This presentation specifically addresses the advantages and limitations of state of the art gel, protein arrays and peptide-based labeling proteomic approaches to assess the effects of a suite of model T4 inhibitors on the thyroid axis of Xenopus laevis.

https://cfpub.epa.gov/si/si_public_record_report.cfm?Lab=NHEERL&dirEntryId=152823