AIM OF THE STUDY

DATASET

IMPORTING THE DATA IN HOUDINI

Visualizing data within this software is certainly not a common thing. However, there are build in features such as the Table Import node which will read CSV files, such as output from spreadsheet applications, and generate a point for each line of the CSV file. Even though it's a faster way to import the data, there was a need to structure those data before importing them. Houdini is build to work with the Python programming language, which is commonly known for numerical and scientific computing.

PYTHON + HOUDINI

Python can be a really powerful tool for structuring data. The first thing to do was to read all the data correctly and then turn them into points. Once the points have been displayed, we can add other attributes to the dataset for visual variables. The points have been transformed into circles, and the radius is equal to the level of expression of each cell. The greater the value of the expression, the larger the circle will be. Each cell selected has its own color which has been configured inside the script. The expression value of the dataset determines the hue of the color based on the value, the higher the value, the darker the hue will be.

https://github.com/DanieleMaselli/Houdini_Python/blob/master/hou.py

CONVERT POINTS TO SPHERICAL COORDINATE SYSTEM

After successfully importing the data, the visual appearance was not as interesting as it was generated with 3D software. Since the x, y, z axis had no great importance on the representation of these data, I used a vex code to transform the points into a Cartesian system. The result was surprising, giving the data an immersive appearance that recalls a galaxy and generative art aesthetics. The source of inspiration as well as the vex code were found on Entagma: https://vimeo.com/229479305

OUTPUT

CONNECTING CELLS

In order to show how the same protein shows different levels of expression in different patients, the lines connects the same proteins but of different cells. This was an attempt to give more sense to the visualization, unfortunately, there was no time to finish this work. The aim of the work was to compare the expression of identical proteins in different patients. For example, the expression of HLADR protein present in the immune cells of healthy subjects in comparison to the levels of the same protein in immune cells of patients (multiple sclerosis).

ARTWORK IMAGES

CONCLUSION

Even though it's a fascinating way to combine two distant world like science and 3D software, we have to take into account the correctness of science as a discipline. This project was a personal experiment and research for a new visual representation of biological data. I am fully aware that this visualization doesn't show an accurate representation of the given study, however I find it still fascinating on how we can manipulate data set with different software that where not build for data visualization. The reaction of the Ph.D student was priceless on how she found the new appereance of her study. This made me think about the natural visual attraction that it generates. Further explanations about visualizing scientific data can be found on the links below, which is a survey paper that i wrote this semester.