Expanding the biological domain: a titration of intelligence into a life-science question

Sweta Rai^1,*

¹Department of Biological Sciences, University of Lethbridge

Lethbridge, Alberta, Canada

Supervisor: Dr. Igor Kovalchuk

^*Correspondence: sweta.rai@uleth.ca

Sweta Rai, corresponding author, holding the Gold Medal — **Plate I.** Corresponding author. The plate is set in binary. The person and the data occupy the same field. She holds the **Gold Medal** of Bangalore University, 2022. Lethbridge, Alberta, Canada.

Abstract

When I learned AI I stayed in biology. I added a titrant. The flask already held a life-science question. How does rheumatoid arthritis write itself into expression and methylation. Computation entered drop by drop. First docking scripts. Then R pipelines for differential expression. Then classifiers that rank biomarkers. I am a PhD candidate in Biological Sciences (Bioinformatics) at the University of Lethbridge. I work with Dr. Igor Kovalchuk on multi-omics biomarker discovery in rheumatoid arthritis. This article is the public record of that titration.

Keywords. Multi-omics integration · Transcriptomics analysis · Rheumatoid arthritis biomarker discovery · Machine learning for biomarker discovery · Molecular docking and cheminformatics · Systems biology

1. Introduction

I began as a biologist. Bangalore University trained me in biology, biotechnology, and bioinformatics. The university ranked me third at the bachelor convocation in 2019. The MSc in Applied Genetics and Bioinformatics ended with the Gold Medal and first rank. The thesis screened roughly three thousand phytochemicals against rheumatoid arthritis targets. That was the first time I added computation to a biological question I already cared about.

I am a PhD candidate in Biological Sciences at the University of Lethbridge. I work with Dr. Igor Kovalchuk on multi-omics data analysis for biomarker discovery in rheumatoid arthritis. I build reproducible R and Python pipelines that turn public genomic archives into clinically meaningful signal from transcriptomics, methylomics, and machine learning.

Alongside the PhD I mentor independent-study and research-assistant students. I help deliver graduate and undergraduate courses in the Kovalchuk lab. I also work on applied data science and AI-evaluation projects in industry. I want to lead a research program across transcriptomics, epigenomics, systems biology, and AI-enabled biomedical discovery.

A flask on the right titrates the career. Figure 2 is a PA radiograph of the left hand. Osteoarthritis on the left film. Rheumatoid arthritis on the right. The PhD is aimed at the second film.

Figure 2. Schematic PA radiographs of the left hand. Osteoarthritis shows osteophytes and asymmetric joint-space narrowing at DIP and PIP. Rheumatoid arthritis shows periarticular osteopenia, uniform MCP narrowing, and radial marginal erosions. Soft tissue is the faint grey envelope. Teaching film. Not a patient study.

2. A formal model of the career

Let $D (t)$ be the domain of the question at time $t$ . The career is an integral.

D (t) = Biology + \int_{2019}^{t} Computation (τ) d τ

(1)

The lower limit is the bachelor year. The integrand is the method the question required next. Docking in 2021. $DESeq2$ and $limma$ in the PhD. Then supervised learners for biomarker ranking. At the equivalence point $t_{eq}$ , intelligence enters. The flask stays full.

Biology + AI t_{eq} D_{expanded}

(2)

Equation (2) is a statement of identity. I still read GEO series. I still argue about batch effect. I still write R. The new term changes the colour of the solution. The same disease question can now be asked with models that were not available at the bench alone.

Figure 1. Titration curve of the investigator, 2019 to 2026. The ordinate is the fraction of each year of work that is computational. The dashed line marks

t_{eq}

, when AI evaluation work at Mercor and the PhD multi-omics programme occupy the same interval. Points are biographical. The sigmoid is a visual aid, in the spirit of an indicator curve.

3. Materials and methods

Differential expression is modelled in the usual way. For gene $i$ ,

lo g_{2} FC_{i} = lo g_{2} x ˉ_{i, RA} - lo g_{2} x ˉ_{i, ctrl}

(3)

with significance from empirical Bayes moderation ( $limma::eBayes$ , $DESeq2$ ) and false-discovery control

q_{i} = FDR (p_{i})

(4)

Transcriptomic and methylomic layers are joined by horizontal, vertical, and gene-level integration. Co-expression uses WGCNA. Causal follow-up uses Mendelian randomization (TwoSampleMR). Feature selection (Boruta, LASSO) precedes classification (logistic regression, random forest, XGBoost). Wet-lab training remains in the reagent list. PCR. Electrophoresis. Nucleic-acid extraction. The full inventory is Table S4 in the methods supplement.

Figure 3. Schematic volcano of a sex-stratified differential-expression contrast. Magenta is up in RA. Green is down. The dashed line is

- lo g_{10} p = 1.3

. Points are simulated. The live analysis is Experiment 1 in Supplement S5.

Figure 4. Schematic expression heatmap for endocannabinoid-system genes, split by sex and RA status. Colour is relative abundance.

4. Results

The flagship academic result is the Gold Medal. The flagship scientific result in print is the 2024 review on $β$ -cell plasticity. The flagship programme now running is the RA multi-omics pipeline. Table 1 records honours. The bibliography records manuscripts.

Table 1. Selected honours and funding

Year	Recognition	Issuer
2022	MSc Gold Medal & University First Rankflagship	Vice Chancellor, Bangalore University
2024 - Present	MITACS Scholarship - Accelerate / Research Training Award	MITACS, with Dr. Igor Kovalchuk (University of Lethbridge)
2019	University Third Rank Holder	Bachelor's Degree Convocation
2023	Certified Data Scientist	Professional Certification
2024 - Present	PhD Scholarship / Graduate Research Funding	University of Lethbridge, Biological Sciences

Table 1. Awarded the Gold Medal and first-rank distinction for the Master of Science in Applied Genetics and Bioinformatics cohort, graduating First Class Exemplary with a CGPA of 8.57/10. MITACS support (2024 to present) funds the doctoral work under Dr. Igor Kovalchuk.

Figure 5. Cartoon PAGE of the career. Each lane is a stage. The Gold Medal is a single heavy band. The PhD lane is a multi-omics smear. AI is a new band at the top of the gel. The bands below stay.

The published review (2024) appears as reference [1]. DOI 10.3390/cimb46070453. A complete list, including manuscripts in preparation, is Supplement S6.

5. Discussion

I am seeking postdoctoral positions in computational biology, bioinformatics, and translational genomics.

I want a laboratory, or a group inside one, where a biological question can pick up a model without putting down the pipette. Letters may be addressed to the corresponding author, or written through Supplement S7.

Acknowledgements. Dr. Igor Kovalchuk for supervision of the doctoral work. MITACS for research training support. Bangalore University for the Gold Medal and the years that filled the flask before Canada.

References

Ghasemi Gojani, E., Rai, S., Norouzkhani, F., Shujat, S., Wang, B., Li, D., Kovalchuk, O., Kovalchuk, I. (2024). Targeting β-cell plasticity: A promising approach for diabetes treatment. Current Issues in Molecular Biology, 46(7), 7621-7667. doi:10.3390/cimb46070453.
Rai, S. (2022). Investigation of anti-arthritis compounds from Cinnamomum zeylanicum through in-silico methods. Master's Thesis, Department of Applied Genetics, Bangalore University.

Supplementary information

The article states the argument. The supplements hold the laboratory book. Vita. Dated entries. Reagent tables. Numbered experiments. The letterhead.