Tag Archive for: bioinformatics

Q&A with Dr. Rokita: Building bioinformatics infrastructure at the Brain Tumor Institute

Jo Lynne Rokita, PhD, is the director of the new Bioinformatics Core housed within the Brain Tumor Institute at Children’s National Hospital.

Jo Lynne Rokita, PhD, is the director of the new Bioinformatics Core housed within the Brain Tumor Institute at Children’s National Hospital. Dr. Rokita is a cancer genomics leader with 20 years of combined research experience in academia, industry and the government. She’s also a technical and analytical expert in genomics research using microarrays and high-throughput sequencing.

“We are very excited that we were able to recruit Dr. Rokita as director of the Bioinformatics Core Facility,” says Muller Fabbri, MD, PhD, associate center director for Cancer and Immunology Research at Children’s National. “Her Bioinformatics Core will play a central role in providing the Brain Tumor Institute community with unique expertise spanning biology/genetics/genomics and bioinformatics and will propel Children’s National forward as a national and worldwide leader in pediatric brain tumor research.”

Dr. Rokita is overseeing the core’s creation, including bringing both bioinformatics staff and computing infrastructure to the program, which will support the data analysis needs of the institute’s investigators. She recently answered questions about the new core and also talked a little bit about the focus of her own research that will continue at Children’s National.

Q: Why is the Brain Tumor Institute establishing a Bioinformatics Core?

A: Growing the institute’s bioinformatics capabilities was one of the things that leadership wanted to make sure was built into the plan for the record-setting $96 million gift that was received in 2023. There was a clear need among the principal investigators for this type of research support which includes organization, analysis and interpretation of large-scale genetic sequencing and other “-omics” data.

Q: How did you decide to join Children’s National?

A: I was leading a pediatric brain tumor focused bioinformatics team at Children’s Hospital of Philadelphia (CHOP). As a part of the Children’s Brain Tumor Network (CBTN), I worked closely with a collaborator from Children’s National, Brian Rood, MD, medical director of the Brain Tumor Institute. He told me about the opportunity and I was very excited to apply.

Q: How did your previous work prepare you for this role?

A: I’ve spent the past 10 years in the pediatric cancer field with the last six focused on brain tumor research. In my various roles at CHOP, I led multiple large-scale genomic analysis efforts, comprehensive data and methods for which we then provided openly to the community. During my postdoctoral fellowship, these efforts included a large neuroblastoma patient-derived cell line “ENCODE” as well as a resource led in collaboration with multiple institutes and the National Cancer Institute funded by Alex’s Lemonade Stand Foundation (ALSF). We further scaled these efforts to build open analytical platforms to empower researchers to build upon our work doing their own cancer genomic analysis. In collaboration with the Childhood Cancer Data Lab at ALSF, we built the platform that ultimately ballooned into the OpenPedCan includes large amounts of harmonized genomic, epigenomic and proteomic data for patients with pediatric cancer. What’s unique is that the data is all processed in the same way and easily accessible through multiple mechanisms. Researchers can use these data to ask questions about the cancer type they study or genes of interest. For example, genes over-expressed, absent and/or mutated in a specific tumor subtype may lead to a better understanding of how a patient’s cancer may respond to a treatment.

We’ll be bringing some of the workflows we created previously here to Children’s National, and that will allow us to join newly generated internal data with the thousands of data points we’ve already harmonized using these workflows.

Q: Can you give us some examples of how data harmonization benefits the field of pediatric brain tumor research?

A: Harmonizing across institutions and databases will help us increase the number of data points available for study. This is really important for rare types of tumors and are major foci of institute collaborator Adriana Fonseca, MD, and her International Rare Brain Tumor Registry program. The Bioinformatics Core will support data organization and analysis for this effort, which aims to sequence the rarest brain tumors — those that make up between only 3% and 5% of all brain tumors. If all the data is analyzed the same way, we can combine multiple studies to increase our total dataset, which in turn may reveal new biomarkers and new subtypes of those tumors. It is critical that we continue to build these data resources in a way that they can be accessed by everyone doing this work. Having dedicated support systems for these functions will push the research farther, faster.

Q: As this work gets underway, what is the core’s main function?

A: As this initiative gets underway, the Bioinformatics Core’s primary goal is to empower investigators by streamlining and centralizing data analyses. We help researchers transfer sequencing data into secure cloud storage, organize newly generated records and prepare those datasets for in-depth study. Our bioinformatics scientists then perform downstream analyses to address the specific questions posed by each investigator. On the backend, we collaborate with information technology at Children’s National to develop a robust infrastructure that supports these activities efficiently. By offering these services in-house, we aim to ensure our investigators have seamless, comprehensive support—ultimately driving innovation and accelerating research progress.

Q: What is “open science” and why is it important in bioinformatics?

A: One of our big focus areas is open science, meaning our goal is to push data and code out into the community so that researchers can easily reproduce and build upon our findings. I’m excited to bring the principles of open science, code sharing and data sharing to the Bioinformatics Core.

Making resources open makes it easier for teams to work together across institutions and research programs. It is also going to benefit patients because people can reuse the code and move towards cures faster. For example, we try to package an entire manuscript’s code when we provide our data so it’s clear how the analyses were done.

Q: What is your particular research passion?

A: I work in several research areas and with many brilliant collaborators. One of our focus areas is understanding how RNA splicing can contribute to pediatric brain tumors to create a change in a protein. We have recently identified tumor-specific splice events in some pediatric brain tumor types and will be partnering with Dalia Haydar, PharmD, PhD, to create therapeutic approaches to targeting these. We are also developing a user-friendly application for mining the large amount of splicing data in pediatric brain tumors.

Another focus of our lab is understanding how the patient’s host genome (alterations inherent in their blood DNA) influences the tumor’s genetics. For example, we’ve just preprinted a study connecting inherited variants to tumor genetics and patient outcomes.

Finally, we are interested in how differences in race, ethnicity and social determinants of health influence survival and treatment outcomes for children with brain tumors.

I am passionate about data sharing, code reproducibility and promoting open science in general.

Q: Is there any specific reason you decided to focus your work around brain tumors and pediatric brain tumors?

A: My cousin passed away from a brain tumor when I was in high school. They didn’t have molecular diagnosis then, but he had a brainstem glioma, likely a diffuse midline glioma. In graduate school, I studied addiction genetics and became fascinated with the brain and towards the end, cancer. As an alumna of Penn State, I was actively involved in philanthropic events raising money for their Dance MaraTHON supporting children with cancer. I was lucky to land a postdoc at CHOP and lean into subsequent roles which allowed my passion for this field to grow.

Q: Last question — What do you do with your time when you are not studying pediatric brain tumor data?

A: I enjoy being with my family, observing my children learn and grow, and listening to music.

Germline microsatellite genotypes differentiate children with medulloblastoma

pastel colored DNA strands

A new study suggests that medulloblastoma-specific germline microsatellite variations mark those at-risk for medulloblastoma development.

Brian Rood, M.D., oncologist and medical director at the Brain Tumor Institute, and Harold “Skip” Garner, Ph.D., associate vice provost for research development at Edward Via College of Osteopathic Medicine, published a report in the Society for Neuro-Oncology’s Neuro-Oncology Journal about using a novel approach to identify specific markers in germline (non-tumor) DNA called microsatellites that can differentiate children who have the brain tumor medulloblastoma (MB) from those who don’t.

“Ultimately, the best way to save children from brain tumors and prevent them from bearing long-term side effects from treatment is to prevent those tumors from occurring in the first place,” says Dr. Rood. “New advancements hold the potential to finally realize the dream of cancer prevention, but we must first identify those children at-risk.”

While analyzing germline sequencing data from a training set of 120 MB subjects and 425 controls, the doctors identified 139 individual microsatellites whose genotypes differ significantly between the groups. Using a genetic algorithm, they were able to construct a subset of 43 microsatellites that distinguish MB subjects from controls with a sensitivity and specificity of 92% and 88% respectively.

“We made discoveries in an untapped part of the human genome, enabled by unique bioinformatics data mining approaches combined with clinical insight,” said Dr. Garner. “Our findings establish new genomic directions that can lead to high accuracy diagnostics for predicting susceptibility to medulloblastoma.”

What the doctors discovered and demonstrated in the study was that MB-specific germline microsatellite variations mark those at risk for MB development and suggest that other mechanisms of cancer predisposition beyond heritable mutations exist for MB.

“This work is the first to demonstrate the ability of specific DNA sequences to differentiate children with cancer from their healthy counterparts,” added Dr. Rood.

Contributing Authors to this research study included:  Brian R. Rood, M.D., Harold R. Garner, Ph.D., Samuel Rivero-Hinojosa, Ph.D., and Nicholas Kinney, Ph.D.

A new algorithm: Using genomics and EHR to detect severe growth disorders

Test tube that says IGF-1 test

Andrew Dauber, M.D., MMSc., a pediatric endocrinologist and the chief of endocrinology at Children’s National, guided research presented at ENDO 2019, the Endocrine Society’s annual meeting, enabling clinicians and researchers to understand the genetic underpinnings of certain pediatric growth disorders, while using electronic health record (EHR) algorithms to screen for presenting symptoms in the exam room. In some cases, this prompts further genetic testing and shortens the diagnostic odyssey for pediatric growth disorders – such as Turner syndrome.

Here is a summary of the research findings, delivered as two oral abstracts and a poster session.

ABSTRACT 1: Presented on Saturday, March 23, at 12:30 p.m. CST

Healthy childhood growth cohort provides insight into PAPPA2 and IGF-1 relationship, revealing a new level of complexity to the biology of growth with implications for the study and treatment of severe growth disorders

Program: Growth, puberty, and insulin action and resistance

Session OR07-5: A Cross-Sectional Study of IGF-I Bioavailability through Childhood: Associations with PAPP-A2 and Anthropometric Data

Background: Insulin-like growth factor 1 (IGF-1) is a hormone essential for human growth and is often bound to IGFBP-3, an IGF binding protein. Pregnancy Associated Plasma Protein-A2 (PAPP-A2) cleaves intact IGFBP-3, freeing IGF-1 to support normal growth functions. This is the first study, led by Dr. Andrew Dauber with collaborators from Cincinnati Children’s Hospital Medical Center, to track PAPP-A2 and intact IGFBP-3 concentrations throughout childhood. The research team studied 838 healthy children, ages 3-18, in the Cincinnati Genomic Control Cohort, to better understand patterns of growth and development by examining the relationship between PAPPA2 and IGF-1 bioavailability.

Study results: Free IGF-1 increased with age. PAPP-A2, a positive modulator of IGF-1 bioavailability, decreased with age, which surprised the researchers, and is not positively associated with absolute levels of free IGF-1. However, higher levels of PAPP-A2 cleave IGFBP-3 resulting in lower levels of intact IGFBP-3, and consequently, increasing the percentage of free to total IGF-1. This demonstrates that PAPP-A2 is a key regulator of IGF-1 bioavailability on a population-wide scale.

Impact: This research may help endocrinologists create unique, targeted treatment for children with PAPPA2 mutations and could help stratify patients with potential risk factors, such as IGF-1 resistance due to increased binding of IGF-1, associated with severe growth and height disorders. See adjoining study below.

Watch: Video interview with Dr. Dauber

ABSTRACT 2: Presented on Saturday, March 23, at 12:45 p.m. CST

Electronic health records can alert physicians to patients who could benefit from genetic testing to identify severe growth disorders

Program: Growth, puberty, and insulin action and resistance

Session OR07-6: Integrating Targeted Bioinformatic Searches of the Electronic Health Records and Genomic Testing Identifies a Molecular Diagnosis in Three Patients with Undiagnosed Short Stature

Background: Despite referrals to pediatric endocrinologists and extensive hormonal analysis, children with short stature due to a genetic cause, may not receive a diagnosis. Electronic health records may help identify patients – based on associated phenotypes and clinical parameters – who could benefit from genetic testing.

Study results: Researchers from three children’s hospitals – Boston Children’s Hospital, Children’s Hospital of Philadelphia and Cincinnati Children’s Hospital Medical Center – gathered data, starting small, with a known variable, or phenotype, associated with severe growth disorders: insulin-like growth factor 1 (IGF-1) resistance. A targeted bioinformatics search of electronic health records led the team to identify 39 eligible patients out of 234 candidates who met the criteria for a possible genetic-linked growth disorder. Participants were included if their height fell below two standard deviations for age and sex and if their IGF-1 levels rose above the 90th percentile. Patients who had a chronic illness, an underlying genetic condition or precocious puberty were excluded. Whole-exome sequencing (WES) was performed on DNA extracted from willing participants, including 10 patients and their immediate family members. The research team identified new genetic causes in three out of 10 patients with severe growth disorders, who were previously missed as having a genetic-linked growth disorder.

Note: Two patients had two novel IGF1R gene variants; a third had a novel CHD2 variant (p. Val540Phe). The two patients with IGF1R variants had a maternally inherited single amino acid deletion (p.Thr28del) and a novel missense variant (p. Val1013Phe).

Impact: Similar EHR algorithms can be replicated to identify pediatric patients at risk for or thought to have other genetic disorders, while expanding genetic research and improving patient care.

Watch: Video interview with Dr. Dauber

POSTER: Presented on Monday, March 25, at 1 p.m. CST

Electronic health record alerts could help detect Turner syndrome, shorten diagnostic odyssey for girls born with a missing or partially-deleted X chromosome

Program: Session P54. Pediatric puberty, ovarian function, transgender medicine and obesity

Poster Board #MON-249: Algorithm-Driven Electronic Health Record Notification Enhances the Detection of Turner Syndrome

Background: Turner syndrome (TS) results from a complete or partial loss of the second X chromosome and affects about one in every 2,500 female births. TS is common in females with unexplained short stature, but the diagnosis is often not made until late childhood (8-9 years), leading to delays in treatment and screening for comorbidities, such as heart conditions, chronic ear infections, vision problems and challenges with non-verbal learning. Using electronic health record (EHR) alarms can help clinicians screen for and diagnose TS patients earlier in life.

Study results: Researchers from Cincinnati Children’s Hospital Medical Center searched EHRs for female patients with idiopathic short stature who met the team’s selection criteria: Their height fell below two standard deviations from the mean for age as well as one standard deviation below the mid-parental height, had a BMI greater than 5 percent and did not have a chronic illness. The search produced 189 patients who met the diagnostic criteria, 72 of whom had not received prior genetic testing. Out of genetic samples available, 37 were compatible for a microarray analysis – which helped the team identify two cases of TS and a third chromosomal abnormality, all of which were missed by routine clinical evaluation.

Impact: DNA samples may not be available for all patients, but clinicians and researchers can identify and integrate tools into EHR’s – creating their own algorithms. An example includes setting up alerts for specific growth parameters, which helps identify and screen patients for TS.

The abstracts Dr. Dauber and his team discuss at ENDO 2019 support ongoing research, including a partnership among four leading children’s hospitals – Children’s National Health System, Boston Children’s Hospital, Children’s Hospital of Philadelphia and Cincinnati Children’s Medical Center – funded by an R01 grant to study how electronic health records can detect and identify novel markers of severe growth disorders.

The researchers hope their findings will also identify and help screen for comorbidities associated with atypical growth patterns, supporting multidisciplinary treatment throughout a child’s life. The study started in August 2018 and includes three sets of unique diagnostic criteria and will analyze WES from dozens of patients over five years.

Read more about Dr. Dauber’s research presented at ENDO 2019 in Endocrine Today and watch his video commentary with Medscape.