Preparing Boys for Life.

11th annual Research Symposium highlights student work in university laboratories

Seven seniors at The Haverford School presented science research projects at the 11th annual Research Symposium. The students, who are part of the Advanced Research Laboratory Cooperative elective at Haverford, completed their research at the University of Pennsylvania, Drexel University, Children’s Hospital of Philadelphia, and other laboratories over the course of six-to-eight weeks in the summer of 2018. The topics of their college-level research included analyzing gene expression data, the possibility of conditioning the brain against severe injuries, the efficacy of a gene therapy used to treat Familial Hypercholesterolemia, and more.

Students spent their fall semester completing formal research papers and presented their scientific findings to a group of their peers, parents, and school administration. Several alumni who took part in the Research Symposium while at Haverford also attended the presentations.

This project allows our students to be the experts in the room. It requires them to develop the process and skills that are required to communicate their research both in a professional scientific paper, as well as in a layperson setting like our symposium. Tom Trocano, Upper School science teacher

The student research papers are submitted to science publications and competitions, and provide the students with a foundation of scientific research skills that can be applied at the collegiate level and beyond. Kara Cleffi leads the program, along with fellow Upper School science teacher Tom Trocano.

“This project allows our students to be the experts in the room,” said Trocano. “It requires them to develop the process and skills that are required to communicate their research both in a professional scientific paper, as well as in a layperson setting like our symposium."

Students apply to take the elective as juniors. Once accepted to the class, students explore research in many scientific fields to determine their topic of interest. Some students will continue their work with the labs after graduation, or during their college studies.

Haverford was able to introduce me to the academic research world. That influenced me to apply to Penn, to pursue graduate studies in neuroscience, and to apply to medical school. Greg Boyek '14

Greg Boyek ’14 is currently pursuing a graduate degree at the University of Pennsylvania, working with the same team of scientists that he worked with as a student at Haverford.

“Haverford was able to introduce me to the academic research world,” said Boyek. “That influenced me to apply to Penn, to pursue graduate studies in neuroscience, and to apply to medical school." Boyek is currently working on a clinical study at Penn’s Translational Neuromodulation Lab, looking for ways to repair the brain using brain-interface technology.

Students are able to expand their scientific background, while also gaining significant exposure and skills that come with working in a real-world setting.

“It was great to work in a professional environment,” said senior Aram LaVan. “Most students don’t get the chance to see this level of work happening. Not only do we get to see it, but we get to be part of it.”

View photos from the Symposium on our SmugMug channel >

Student project abstracts are below:

Nicholas Biddle 
Understanding How Endocannabinoid Pathways Effect Sleep and Seizures in Drosophilae Melanogaster
The Sehgal Lab at the University of Pennsylvania

Endocannabinoids and their relation to sleep and seizures remain uncertain. Endocannabinoids are similar to the plant-based cannabinoids in both structure and the receptors they interact with. Cannabinoids are used to treat those with insomnia as well as those with epilepsy, so endocannabinoids might potentially do the same. To determine if and how endocannabinoids affect sleep, assays using drosophila melanogaster given Arachidonoyl Taurine (AT) and Arachidonoyl Acid (AA) as well as qPCR of CG8839, an endocannabinoid-equivalent enzyme, in drosophila were performed. To further understand what channel methanandamide, an endocannabinoid that reduces seizures in drosophila, is an agonist for when reducing the number of seizures in drosophila, assays using drosophila given AMG9810, Capsazepine (CPZ), and the mutation painless were performed. 

The assays for AT and AA found that drosophila given AT had increase sleep duration while those given AA had decreased sleep. These results support that endocannabinoid pathways are linked to sleep and seizures. QPCR of CG8839 resulted in decreased expression of flies at ZT12 and increased expression of those sleep deprived. These findings, however, suggest that CG9939 pathways do not operate as previously thought, but that cannot be substantiated completely by these results. 

The assays for AMG9810 and CPZ found that both of those TRPV1 inhibitors increased the amount of seizures in drosophila given methanandamide (mAEA), but those with the painless mutation and given methanandamide had close to zero seizures. These findings suggest that the TRPV (Transient Receptor Potential Subfamily V) channel, TRPV1, might be what mAEA operates through, but the results cannot support that completely. 
Nikhil Chakraborty 

Analyzing Demographics of EAT Cafe, a Pay-What-You-Can Cafe, to Determine Groups of Need Served
The Center for Hunger-Free Communities in the Dornsife School of Health at Drexel University

Pay-what-you-can (PWYC) pricing is a model where customers pay what they can
afford. This paper analyzes the demographics of one PWYC cafe in West Philadelphia: EAT
Cafe. Servers collected data on consumers’ age, gender, race, and whether they came with their
families through observational tallies. These data were compared to the amount paid and how
often the customer came, since paying habits and repeat customer volume were assumed to be
indicators of need. Results showed that people of color, adults aged 18 - 40, and single males
were groups that needed and used the Cafe the most. This information will be useful for future
outreach efforts and applying for targeted grants.

Nick Chimicles 
Frequent Item Set Analysis of Breast Cancer Gene Expression and Copy Number Data Unveils Novel Gene Interactions
The Children’s Hospital of Philadelphia

Past research has applied data mining techniques to clinical breast cancer data, but never
to genetic data. With pharmacogenetics being the future of breast cancer treatment, genetic data
must be analyzed to uncover genetic targets. Targeting specific oncogenes could undermine

In this study, we applied Frequent Item Set Analysis (FRISCO) using the programming
language R and the “apriori” algorithm to breast cancer expression and copy number (CNA) data
from The Cancer Genome Atlas (TCGA). We hoped to discover networks of genes that were
co-overexpressed or co- copy number amplified. A Kolmogorov–Smirnov test was used to filter
only unusually expressed genes . We used a percentile analysis to categorize the expression data
into over and normally expressed. The CNA values were separated as either over or normally
amplified. We then merged the two transaction tables of expression and CNA data. An AWS
machine was employed to feed this data into the apriori algorithm. The output of this algorithm
was analyzed using heat maps and a cytoscape network (figure 1).

We identified three main groupings: one with ERBB2, another with FADD and a third
with genes from the 8p12 and 11q13 regions. HER2 is a well-known indicator of a specific type
of breast cancer. FADD has been identified as involved in breast cancer, but not extensively
studied. The 8p12/11q13 genes are scantily understood in their relation to breast cancer.
We hope that the genes uncovered in this study will be further researched for eventual
targeting and that FRISCO will be used for analyzing the genetic data of all cancers.

Mickey Fairorth 
Construction of a Plasmid Vector Containing SLC39A8 Variants in Order to Test the Effects of ZIP8 Mutations in a Cell Line
The University of Pennsylvania 

Previous studies have shown an important role for the metal ion transporter ZIP8,
encoded by the SLC39A8 gene, in Mn homeostasis. 1,2,3,4,5 When mutations in ZIP8 occur, Mn
levels become deficient which in turn causes Congenital Disorders of Glycosylation (CDG) and
Leigh-like Syndrome. 2,3,4,5,6 Our experiment, which looked to examine the specific functionality
of ZIP8 and the effects of two gene variants, relied on a plasmid containing the entire SLC39A8
gene. The two variants, P249A and A391T, were inserted separately into plasmids and
transformed into E.coli bacteria. No bacterial growth was observed, implying issues with
plasmid assembly or with the bacteria themselves. The next step in this process would be to
Sanger Sequence the plasmid backbone as well as the two variant inserts to ensure the proper
sequences and then repeat our procedures.

Aram LaVan
LDL clearance from Plasma in a Patient Treated for Familial Hypercholesterolemia with an AAV8.TBG.hLDLR Vector
The Perelman School of Medicine, University of Pennsylvania

Familial Hypercholesterolemia (FH) is a disease caused by faulty receptors for LDL in
the liver. LDL carries cholesterol to the tissues and is metabolized using the receptors, so the
lack of proper receptors vastly increases the cholesterol level in the body. This can cause
early-onset and severe cardiovascular disease as well as cholesterol deposits in the skin, tissues,
and around the eyes. In the Smilow Center for Translational Research, the Rader Laboratory
looked into the efficacy of a gene therapy used to treat FH. Labeled ApoB-100 protein on LDL
molecules were measured after an infusion of D3-leucine. The level of leucine was expected to
increase during the infusion and decrease after. However, the level did not increase as
significantly as was predicted and did not lower after the infusion had stopped. Additional LDL
samples must be examined to fully understand the reasons behind these results. If, however, gene
therapy proves to be a successful procedure for lowering LDL in FH patients, it could vastly
increase the quality of life of those afflicted with the disease.
Neetish Sharma 

Visualizing Microglia Activation to Investigate Mild Traumatic Injury Prerecording (mTIP) in Mus musculus Brain Models
The University of Pennsylvania

While the majority of studies indicate a compounding effect of multiple mild brain injuries,
reports have suggested that a prior insult can “condition” the brain against future ones, a
phenomenon referred to in this study as Mild Traumatic Injury Preconditioning (mTIP). The
underlying purpose of this study is to gain a greater understanding of the role of immunoprotective
microglia in the mTIP effect. Specifically, microglial activation “conditions” the brain against
more severe injuries. To visualize microglia activation, mice were given controlled cortical
impacts (CCI) to model different injury conditions, and brains were perfused, cryoprotected, and
frozen to prepare for histology. After brains were cryosectioned into slices, immunohistochemistry
was performed to label microglia with IBA-1 antibody; fluorescence images were then taken and
processed via iterative pixel threshold segmentation for IBA-1 percent area, which represented
microglia activation within certain cortical and hippocampal regions. Findings indicate that the
mTIP phenomenon was exhibited structurally via the activation of microglia within specific
regions of the brain, depending on the injury conditions. Essentially, it was demonstrated that a
prior injury leads to microglia activation that observes the mTIP effect, preconditioning the brain
against future insults. Thus, this study supports the idea that microglia are an accurate biomarker
for the level of impairment associated with a given CCI. Adhering to the mTIP effect, microglia
activation increases and decreases accordingly with the level of impairment for each injury

Will Vauclain - A Novel Approach to Gene Expression Databases
Childhood Cancer Data Lab sponsored by Alex’s Lemonade Stand Foundation

In the field of genomics, analyzing gene expression data is becoming increasingly
important. Since data often comes from multiple different platforms, making analysis difficult,
there is a clear need for data homogeneity. The goal of this study was to develop a database for
gene expression data that is more homogenous than existing sources. To do this, we developed a
pipeline that obtains microarray and RNA-Seq data from three publicly available sources and
converts the metadata and gene expression files into a common format for storage. These data
were then made publicly available on a website hosted at “”, which allows researchers
to create and download datasets comprised of multiple different studies for their own analysis.