Blog #2

We collected all the raw data from Yahoo and combined all the data into one data set.

Firstly,we tested the temporal Taylor’s law by group the data set by company and calculate mean and variance for each one of them so we could evaluate each stock’s performance since they went public. We also built a linear regression model for the mean and variance on a log-log scale to see how the strong the relationship.

This is the mean-variance plot we got for temporal on a log-log scale and we can see it is a linear relationship which matches with Taylor’s law and we can also clearly see some colored outliers which either has significantly high stock price or relatively low stock price. We also test the linear regression model the R-square value which is use to evaluate the strength of the relationship is at around 66% indicates a relatively strong relationship exists between the mean and variance for each stock.

Then we tested the ensemble Taylor’s law which we grouped the data set by year and evaluate the market performance on a yearly basis. From this graph we can observe one interesting thing is that there is push back on year 2008 which is the year when financial crisis happened and we can also see the stock price has been recovering after year 2008 which show us that this Taylor’s law is able to reflect the overall market performance. The strength of this model is 67%.

The last one we tested is the temporal hierarchical Taylor’s law. For this one we will have to group the data set by year and company. Therefore we would be getting the information about how each stock is doing each year and calculate mean and variance of each year for every stock. To clearly see how each stock is doing we built a 5×5 multi-panel and we can see most of the relationship is linear. And the strength of the model is much stronger since there are more variables come into play, the R square is 87%.TH

Blog Post 2: The Effects of Peer Mentoring on First-Year Student Athletes

For the project, I have needed help both from the faculty member and others working in the tutoring center with me. The biggest issue I faced was ensuring that all the Peer Mentors were inputting data uniformly. Initially, some were using notations different from the structured format. Another less consequential issue was collecting data from students who had missed their mentoring session.  Mentoring sessions were conducted weekly, hence collecting missing data did not require finding or searching for the participants. Fortunately, the mentors were able to meet with the students weekly so if they missed one week then they were able to gather data the following week. The data validation was based on errors found; there is a structured format for the data sheet where we collect the data, thus there is a finite amount of answers that can be given for each of the questions posed. Other than this, the collection of data has been done rather smoothly. During each session the mentors conduct with the mentees, they incorporate asking the quantitative question into the lesson. Early on, we discovered the most challenging aspect was determining the overall effectiveness of the mentoring program.  We relied on the self-reporting of the students who participated in the program. As with all self-reporting, the results were subjective and hardly quantifiable. We sought another method to see if there were any way to quantify how the strategies we were sharing had an impact on the academic success of the participants. We determined eight areas where we could use objective data to see how much time students used to employ the strategies for academic success.

Every Wednesday I meet with Professor Buffone and we discuss how the research is being collected, if there are any discrepancies in the data, what kind of data we should be taking moving forward, and also what I have been noticing when observing the data. Throughout the week, he checks in with me if he finds an issue with the collection of data or if he thinks I should be working on certain parts of the research more than others. Professor Buffone keeps me motivated to stay on track and timely with the work needing to be done because he is the one who sets the deadlines and holds me accountable for the work I do and for meeting benchmarks. 

We have found that there might be a correlation between 100% attendance and 0% attendance; those with either of these traits might have higher GPAs, but we have to wait until after the semester is over to test this theory. According to this theory, some students might have come in with very competent skills, thus they are independent learners without the need for assistance from any outside parties. We have discussed what can affect GPA and when we come back from winter break, we will be able to draw conclusions from the data.  We have already discussed various methods of analyzing the data that can be used to extract conclusions. 

We have asked how the data collected affects the GPA of the student-athletes. Since we have conducted a preliminary review of last fall’s data, we have been able to preview what types of outcomes we expect to find after this semester. Some speculation on an outcome has been produced from the preview. Unfortunately, we do not have both semesters to base our outcomes on yet, so we are unable of what type of conclusion we are expected to find.

Blog 2: Presenting my research at the American Speech- Language-Hearing Association (ASHA)

Attending the American Speech- Language-Hearing Association (ASHA) Convention was a meaningful and extremely worthwhile experience as an aspiring Speech-Language Pathologist (SLP). An estimated 20,000 people of all different professions united with two interests alike, speech and hearing. Dr. Gregory and I presented our poster “Cultural Humility: Examining Microaggressions to Improve Clinical Encounters” on the first day of the three-day convention. Our poster was displayed amongst hundreds of researchers work on various topics including telepractice, cognitive disorders, health literacy, language in infants, and others focusing on different aspects within our career path. During our presentation time, professions from different regions of the world shared their perspective on microaggressions in the workplace and everyday life.

We were apart of relevant and sincere conversations that taught us new things regarding personal bias’ while being able to provide research that benefits all individuals within a community. A key take away from our presentation was the amount of people who identified with “self- evaluation” posing a huge barrier to achieving cultural humility. Self- evaluation is the ability to reevaluate and alter personal biases with the willingness to explore and appreciate a culture for what it is. Many people we spoke with addressed microaggressions as a “sensitive topic,” which placed even more emphasis on the importance of talking about how they impact our clients.

As we take all considerations and critiques away from our experience at ASHA, the next step is to implement our poster and checklist into local university clinics.  Additionally, we would like this checklist to be feasible for practicing clinicians and professors to introduce in their coursework.  As we try to spread awareness on the impact of microaggressive attitudes on a national level, it is equally important to ensure that we enforce the same beliefs here in our community. While continuing to focus on cultural humility, the survey and qualitative interviews on cultural competence and humility are underway as we work toward building questions that will give us reliable and significant results. The aim of this study is to identify undergraduate and graduate student experiences with microaggressions during clinical experiences.



Blog 2 Environmental Discourse and Diversity in Urban Setting

This semester, our team has discussed the problems and difficulties of the project. First, we talked about how to gain access to the interview with potential research participants. My teammates shared their stories and tips; as for me personally, I found asking permission from individuals was much easier than asking the institutions directly. In the meantime, we realized that maintaining a good relationship with the participant after the interview had been done was equally important. For example, when I completed my interview, the interviewee invited me to join the opening day of her businesses, which on the other hand could also be an opportunity for further research studies in the future. Finally, we also talked about how to overcome our shyness when we reached out to the participants, creating a comfortable space for them as well as us.

The discovery of the horticultural community in New York City has been an exciting study for me on the project. I learned how they developed and practiced their philosophy in this small subculture; having plants in their lives meant maintaining self-care in an urban setting where the power of nature was strictly limited. Moreover, the fast-paced urban life also had an impact on their practice of urban horticulture. For example, many owners had to abandon their plants while moving to new houses. Thus, the community developed the idea of “rescuing plants” or “plant adoption”, similar to animal rescues, to share the responsibility of taking care of the environment.

Compared to my teammates who focus on nonprofit environmental-related institutions, I decided to concentrate on the business aspect of environmental discourse in New York City. I have discovered a few businesses which try to maintain environmentally ethical while gaining profits. I hope I will expand the research data and include more research participants in future studies.

Blog 2: Effect of Diet of Microglial Dynamics in PTEN-ASD: Wildtype Fish Preliminary Results

For my project so far I have analyzed microglia dynamics of wildtype fish that were fed 3 different diets (high fat, high glucose, and high protein (control). So far I have found differences between microglia dynamics (area and movement) of wildtype fish fed a high fat diet compared to the other two diets. I have spoken with my faculty mentor about the possible explanations behind these observed differences and we have came up with a few hypotheses that we are planning to test next semester. I have also been looking at scientific literature about microglia motility and the effects of a high fat diet on microglia as well, to help explain my preliminary results. I have also presented my preliminary findings at the monthly biology research group meeting we have here at Pace and I received some input from other professors on what we could do to get a better understanding of what our results mean. 

My results of my experiment have found that the microglia of wildtype fish fed a high fat diet had a significantly larger cell size as well as faster velocity. Additionally, I have found that microglia of fish in this condition moved greater distances and also moved for less time overall. Other research done on the effects of high fat diet have also found signs of microglial dystrophy, one of them being soma enlargement. While this helps validify my findings, I plan to analyze more fish (my current number of fish analyzed is 6) to see if this trend holds up. Interestingly, the research conducted on the effects of high fat diet on microglia did not analyze motility, which makes it difficult to verify my findings. However, I found that there was some research conducted on epilepsy’s effect on microglial motility in a mouse model, which could help give me some basis to compare my results, as epilepsy is characterized by neuroinflammation and microglia activation, which is similar to the effects of high fat diet on the brain. 

This study found that microglia in epilepsy model mice moved greater distances, while maintaining their velocity. This partially matches my findings, as I also found that microglia in the high-fat diet condition moved overall greater distances, however I also found that they moved at higher velocities. Again, I plan on analyzing more fish to see if this trend continues. Overall, after speaking my faculty mentor, we devised a few more tests to get a greater grasp on what are the effects of diet on microglia. 

Some of these include acridine orange staining (will help us analyze neuronal death to see if microglia are reacting to neuronal damage induced by certain diets), sholl analysis (will help us analyze differences microglial morphology in response to certain diets), annexin-v staining (will help us see if microglia are undergoing phagocytosis), cell counts (will help us see if microglia are proliferating in response to certain diets) and western blots (will help us see if microglia are producing pro-inflammatory cytokines in response to certain diets). Additionally, next semester we are getting a new microscope in the lab, which will help us get a total view of the zebrafish tectum, allowing us to analyze all of the microglia within it. 

In closing, my current hypothesis based on my results is that a high fat diet causes increased microglial activation, which can be evidenced by their enlarged cell bodies, and lower duration of movement, as both imply engulfment of cellular matter, which microglia are known to do in their activated state. However, further analysis using the methods I described above will give me more of a clue if this is what is indeed occurring.


BLOG 2: The Effect of the Combined Treatments of the CDK4/6 inhibitor Palbociclib and the ACL inhibitor Bempedoic Acid on Apoptosis in Breast and Pancreatic Cancer Cells.

From previous experiments, we found that the combined treatment of Palbociclib and Bempedoic Acid would lead to the decreases of cancer cell numbers. In order to figure out the pathway that causes the cell death, we have conducted more experiments on breast cancer cells 231 and pancreatic cancer cell Panc1 since the beginning of this semester.

We first assumed that these cancer cells underwent apoptosis after the combined treatment. Caspase-Glo® 3/7 Assay was used to determine the apoptotic process of 231 cells in a standard 96-well assay plate. On the first day of the experiment, 6,000 cells were counted and placed in each well from well B2 to E7 in the plate. On the second day, we added Media, DMSO, Palbociclib (P), Bempedoic Acid (B), P+B, Staurosporin (ST) to columns 2 to 6 respectively, four trials for each drug (B:E). Staurosposin is a drug that triggers apoptosis in cells. Four days later, we observed that there were fewer cells in the P+B group and almost no cells in the ST group. After we made Caspase Glo reagent, added them to the place and read the plate, the fluorescence level of the cells indicates that most cell death in the ST group was caused by Apoptosis, whereas in the P+B group, not many cells underwent apoptosis. We repeated this experiment once and obtained the same results. Because the plate were read 3 days after we added the drug, the apoptotic process might had occurred between Day 2 and Day 5 so that we failed to catch it. Therefore, we decided to use the RealTime-Glo® Assay in our next experiment. It allowed us to measure apoptosis in the plates multiple times for days without killing the cells.  Meanwhile, we wanted to test if the cell number decrease under the combined treatment is caused by other pathways like ferroptosis.

Ferroptosis is a non-apoptotic, iron-dependent, oxidative cell death primarily implicated in inflammation (Skouta et al., 2014) Accumulation of reactive oxygen species (ROS) in a cell can lead to degradation of cell membrane and other damages that are lethal to the cell. The cell can detect such damages and trigger ferroptotis. It has been found to be involved in diseases like cancer, myocardial infarction and many neurological diseases (Li et al., 2019). Researches done by Skouta et al. indicate that ferrostatin is a potent inhibitor of Ferroptosis (2014). Ferrostatin (F) inhibits cell death by processing through a reductive mechanism to prevent damage to membrane lipids (Skouta et al., 2014). Our second assumption is that the cell number decrease is caused by Ferroptosis. If our assumption is right, we should observe cell death rescue when we add ferrostain to the combined treatment.

In our fifth experiment, Panc1 cells were placed in the standard 96-well assay plate (3000 cells/well, B2:E9). On the second day, we added Media, DMSO, P, B into Columns 2 to 5 respectively, and P+B into Columns 6 to 9. In Column’s 7 to 9, ferrostatin of different concentrations (10µM, 1µM, 100nm respectively) were added. From Day 3 to Day 6, the cell number change was observed under the microscope and any apoptosis was observed using the RealTime-Glo® assay and reagent. We did not detect significant apoptosis in the P+B group. Thus, it was drawn that the cell number decrease was not caused by apoptosis, and we focused on detecting ferroptosis rather than apoptosis in the following experiments. In the meantime, ferrostatin did not have a significant effect on rescuing cell death. We believed that the Panc1 cells we used were old and did not actively react to the drugs. Also, the ferrostatin concentration might be so high that it caused cell death.

In Experiment 6, new Panc1 cells (3000 cells/well) were placed in a new plate (B2:E11). On the second day, we added Media, DMSO, P, B into Columns 2 to 5 respectively, and P+B into Columns 6 to 8. In Columns 7 and 8, ferrostatin of different concentrations (10µM, 1µM respectively) were added. In Columns 10 and 11, only ferrostatin was added to the cells (10µM, 1µM respectively). On Day 6, the Cell-Titer Fluor Assay was used to read the plate, which measures the relative number of viable cells in culture. As a result, we found that ferrostatin (Column 10 & 11) did not lead to cell death, meaning it is safe to use ferrostatin as the ferroptosis inhibitor. In the meantime, ferrostatin still did not show signs of rescuing cell death.

In Experiment 7, we counted and plated Panc 1 cells using the same data from Experiment 6. On the second day, we placed the same drugs into Columns 2-5 and added P+B into Columns 6-11. In Columns 10 and 11, ferrostatin of different concentrations (10µM, 1µM respectively) were added. In Columns 7 and 8, we added N-acetyl-L-cysteine (NAC, 1µL & 2µL respectively) to the P+B treatment. NAC is a ROS inhibitor. If the Panc1 cells did undergo ferroptosis, NAC would prevent ROS accumulation from signaling the cell death, which means the number of viable cells in these two columns will be higher than that in the P+B group. Cell-Titer Fluor was used to detect the numbers of viable cells in each column. Our results from Experiment 7 suggested that there wasn’t enough cell death in the P+B group. We decided to decrease the cell confluence in each well by placing 1000 cells/well in Experiment 8 while keeping other data the same. Results from Experiment 8 indicate that the cell number in the P+B group significantly decreased, and the cell death was rescued in the P+B+1µM F group. There was no significant cell number increase in the P+B+NAC group. We will repeat more experiments to compare cell numbers in the P+B and P+B+1µM F group in order to draw our conclusion that the cells underdo ferroptosis. We found that the cell numbers in the wells at the corner of the plate are relatively lower than those in other wells. We believed that it resulted from evaporation. Therefore, in the following experiments, we will add 200µL phosphate buffered saline (PBS), a salt solution used to wash and transport cells, around the wells we plated cells in, to minimize the effect of evaporation.

These are the data and results we have obtained so far. We have found that the cell number decrease under P+B combined treatment did not result from apoptosis but was very likely caused by ferroptosis. In future experiments, we need to manipulate the concentration of P & B in the treatment in order to get a larger decrease in cell number. Also, we plan to do the same experiment on breast cancer cell 231 to see if ferrostatin is able to rescue cell death in different cell types.


Li, X., Duan, L., Yuan, S., Zhuang, X., Qiao, T., & He, J. (2019). Ferroptosis inhibitor alleviates Radiation-induced lung fibrosis (RILF) via down-regulation of TGF-β1. Journal of Inflammation, 16:11.

Skouta, R., Dixon, S. J., Wang, J., Dunn, D. E., Orman, M., Shimada, K., … Stockwell, B. R. (2014). Ferrostatins Inhibit Oxidative Lipid Damage and Cell Death in Diverse Disease Models. Journal of the American Chemical Society, 136(12), 4551-4556.

Blog 2 – Using Charmm-GUI: Combining Docking, Molecular Dynamics Simulation, and Free Energy Calculation to Improve the Accuracy of Virtual Screening of Anticancer Drug Candidates Against Human Telomeric G-quadruplex DNA

Previous issues that Dr. Deng’s research has encountered include inaccurate 3D visualization of results. This issue is not due to incorrect input or human error, but rather due to the limited abilities/extent of the algorithm in molecular docking programs previously used. To solve this issue, we are currently trying to utilize a different molecular docking program called Charmm-GUI. Charmm-GUI is designed to “interactively build complex systems and prepare their inputs with well-established and reproducible simulation protocols for state-of-the-art biomolecular simulations” (as described by its official website). In coordination with widely used simulation packages such as CHARMM, AMBER, NAMD, GROMACS, GENESIS, LAMMPS, Desmond, and OpenMM, Charmm-GUI is a very promising platform whose more streamlined interface will hopefully provide more accurate results for our virtual screening of anticancer drug candidates against human telomeric G-quadruplex DNA.

Work for this project under Dr. Deng has been going smoothly so far. Regular communication has been crucial, and I am fortunate that Dr. Deng is always available to provide insight and tips in the learning process for this new molecular docking program and the overall problem-solving process. We plan on continuing research during Winter Break going into the Spring Semester, during which we hope to obtain final results.

Blog 2: A continuation in the study of Anonymous Dynamic Distributed Computing without a Unique Leader

Since the last blog the progress that me and my professor have been making so far in strategizing ways of running our simulations in a faster way to collect as much data as possible.

Some of the methods we have discussed have mainly involved paralleling the section of code involved in matrix and vector multiplication in which an adjacency matrix is multiplied by a vector to give us our result in terms of rounds of phases for each node.

The work so far has been very intensive in not only understanding very deeply the Java library of multithreading but also the process of parallelism within a computer. How there is a significant difference between working with a sequential machine and a parallel machine and how the difference has a significant hinderance on the progress of the simulations. Because of that I have been studying very intensely fundamental operating systems concepts of context switching ,proper thread synchronization,  managing multiple processes within a single thread, fork-joins-pool methods etc.

This has lead to good personal experience by learning that theory and practice are two very different things. Theoretically with proper multithreading we were able to figure out that our work could be speed up from O(n^3)  computational time to O(lg(n)) computational time. To put that into perspective something something that takes 1000 days in O(n^3) could take only one day with O(lg(n)).That’s a difference of one day to nearly 3 years!However in practice by having more threads we are creating more work to be done with each thread with parallel programming and thus that could lead to the creation of more time if done recursively enough and especially if it exceeds the number of available cores we have which before used to be only 12 cores; no more than an average laptop!

For this reason we have been working intensely on speeding things up as much as possible in computational time and have been working with .slurm files so that we can tell that machine not only the files we want to run but also the commands and directives in terms of how long we want the simulations to run and how many tasks/CPU cores each task should run which I will add has been a very nice learning experience for myself running the batch commands on a very powerful school computer that is primarily done for research purposes.

Also I have learned to regularly meet with the professor as I do research and have first hand understood the dynamics of student-faculty research by communicating very clearly and promptly with my professor about issues or concerns with running the simulations or the amount of time that needs to be accounted for an all the meta-engineering tasks/problems that are involved with writing a theoretical paper.

For now we are both highly motivated to gather the data/run simulations and study it as much as possible now that we both know we currently have access to a  G.P.U machine with 3,840 cores available for our work we look forward for the work ahead and the research to come.


Blog #2

Over the course of the last several weeks, my faculty advisor (Dr. Kristen Di Gennaro) and I have been working on creating a survey in Qualtrics. Currently, we are waiting on IRB to get back to us on account of the quality of our survey.

Our survey is designed to discover some of the consequences of the linguistic reforms and determine whether they have been successful. One of our questions is if participants have an awareness of the prescriptive grammar rules. After the surveys approved and completed, the next step would be to determine the reasons why some might be more inclined to change their speech patterns, while others remain more conservative when it comes to speech. The purpose is to find out if the adaptability to new rules is related to age, gender, educational level or even political views.

The survey has four parts, that is not counting the disclaimer for voluntary participation. Each section has four parts. The first block is focused on finding out what kinds of pronouns would a participant use when they have a choice. That is, they are asked to rate a sentence on a scale from “I’m very likely to use this phrase” to “I’m not at all likely to use this phrase”, with two more possible answers in between. The purpose is to find out how particular participants might be when it comes to the prescriptive grammar of pronoun usage.

The second part of the survey is designed to see whether participants are more likely to use gender-neutral terms in relation to professions or not. The third section would appear somewhat ambiguous to some participants since unless one is intensely familiar with prescriptive grammar rules (for example, a sentence can’t end with a preposition), the sentences would appear correct. This section is specifically aimed at getting the information on how much of a hold the prescriptivism has on the minds of participants.  The fourth section is fill-the-blank type, and is also aimed at the pronoun usage; however, unlike the first section, it is letting the participant input their own response in order to see if they would naturally put “they” instead of “he” or “she”. The questions are purposefully ambiguous for that reason.

Overall, we anticipate the survey to reveal thinking patterns that relate to prescriptive grammar usage. We are curious if people of older generations are less prone to adopting linguistics reforms. Yet, it must be mentioned, that one of the articles (Mucchi-Faina, Gender identity, and power inequality) that we used to preliminary research states otherwise.

Blog 2: The Effects of Positive Emotion Overexpression Based on Underlying Motives

For this research project, I and Dr.Gosnell have made monumental progress. That is, completing our IRB and sending it in for approval to receive permission to run the study. As for the problem-solving process, I was challenged in trying to learn the format for an IRB and how to find reputable measures to include in the IRB to receive approval. Dr. Gosnell did grant me a very generous amount of initiative in creating  the IRB and gathering the necessary resources for it. Therefore,  I do believe the problem-solving process, the initiative needed for this project, the communication and teamwork/collaboration with my faculty member all go hand-in-hand. This is because each Thursday Professor Gosnell and I would meet to discuss revisions for the IRB, what it had enough if, what it is lacking, and the educational resources I need to make it better. Therefore, Dr. Gosnell acted as a significant problem solver in the behind the scenes work that went into making the IRB, as well as allowed for many initiatives in granting me the freedom to draft the IRB up.

As for insight and reflection into the data I obtained, it all stems from gathering the different types of questionnaire scales needed in the IRB. I used Google Scholar and PSYCHINFO for a lot of the data I obtained, for which I then learned how to decipher if a scale is reputable and valid to use for research purposes. From the data I collected, the only questions that I have raised would be how exactly are the scales we chose going to play out in the facilitation of our research, and how are our participants going to respond to them? Will they be truthful, deceitful, etc.?