Hi, I’m Raul Duran de Alba, a biotechnology engineer and bioinformatician interested in connecting experimental biology with computational models. My training has taken me from biotechnology at Tecnológico de Monterrey, Guadalajara, to the GENIOMHE master’s program in Bioinformatics at Université Paris-Saclay, which I completed in 2026.
I enjoy both sides of biological research: working with cells and experimental systems, and writing code to understand what the data can tell us. The questions that interest me most involve how cells change state, how those changes contribute to disease, and how we can design experiments and models that inform one another.
My research background
For my master’s internship at IBENS, I worked on Boolean gene regulatory network models of focal segmental glomerulosclerosis (FSGS), a kidney disease. The project focused on podocytes and parietal epithelial cells. My work explored how RNA-seq preprocessing choices and differences between datasets affect the ability to infer and compare models. It made me especially interested in a practical question: how can we tell whether data generated in different laboratories are compatible before building a model from them?
At I2BC, I worked on a Snakemake workflow for SLAM-seq and RNA kinetics in yeast, using R-based analysis tools. Earlier, as a research trainee at Harvard Medical School / Brigham and Women’s Hospital, I gained experience with vessel-on-chip systems, cell culture, and expansion microscopy. I also participated in iGEM 2019, working on the Gliksys bio-ink project.
Across these experiences, I have worked with Python, R, reproducible analysis workflows, RNA-seq, and biological network modelling, alongside experimental biology. I am particularly interested in bringing mechanistic models together with machine learning, tissue engineering, and experimental design.
Why I am taking HTBAA
I want to move between understanding biological systems and building with them. How To Biomanufacture Almost Anything is an opportunity to connect my computational and experimental interests with automation, biological design, and the practical constraints of manufacturing.
My first course proposal explores a probiotic bacterial community for vitamin D production and absorption, motivated partly by nutritional challenges in Mexico. It is an ambitious research idea: producing vitamin D inside the gut without light remains unresolved, and improved absorption would need to be demonstrated.
Outside science, I am interested in art history, contemporary art, and curatorial practice.
This site documents my assignments, project ideas, and learning throughout HTBAA — Fall 2026. Use the navigation bar on the left to explore my work.
Week 1 — Biomanufacturing project idea
A Living Probiotic Community for Vitamin D Production and Absorption
A proposal for a two-member bacterial community, including candidate biological roles, an automation strategy, manufacturing estimates, and the main technical unknowns.
Subsections of Raul DURAN DE ALBA — HTBAA Fall 2026
Week 1 HW: Final Project Ideas
A Living Probiotic Community for Vitamin D Production and Absorption
Raul Duran De Alba — HTBAA Homework #1 — September 22, 2026
My proposal is a bacterial community designed to produce a small, controlled amount of vitamin D inside the gut and improve its absorption. This is an ambitious research concept: in-gut vitamin D production is an unresolved challenge, and the proposed functions must be distinguished from what has already been demonstrated.
1. What would I like to biomanufacture, and why?
I would like to biomanufacture a defined community of probiotic bacteria with two complementary functions: producing a small, controlled amount of vitamin D inside the intestine and helping the body absorb it. The final product would be a capsule containing selected bacterial strains, each assigned a specific role.
My motivation is an interest in personalized supplementation: using living systems to help address nutritional needs while accounting for differences between people’s microbiomes. I am particularly interested in its potential relevance to Mexico. An analysis of ENSANUT 2018–19 reported vitamin D deficiency in 31.6% and insufficiency in 46.1% of Mexican women aged 20–49, using the study’s definitions. These results identify a substantial problem in that population, although they should not be generalized to all Mexicans. Source: Contreras-Manzano et al., 2021.
For an initial engineering target, I would aim for 5 micrograms of vitamin D produced per person per day, equivalent to 200 IU. This is approximately one-third of the recommended daily intake of 15 micrograms for adults aged 19–70. It is a proposed production target, not a treatment dose, and the amount absorbed would need to be measured separately. Source: NIH vitamin D fact sheet.
2. What living system would I use?
I would begin with a two-member bacterial community:
Community member
Proposed role
Evidence or assumption
An engineered strain of Bifidobacterium longum
Candidate chassis for producing and releasing vitamin D
This is a hypothetical engineering role. A suitable production pathway has not been established for this proposal.
Limosilactobacillus reuteri NCIMB 30242, formerly called Lactobacillus reuteri
Candidate partner for improving vitamin D availability through interactions with intestinal and bile-acid metabolism
Human evidence motivates testing this strain, but does not establish the mechanism or guarantee improved absorption.
A post hoc analysis of a randomized trial involving 127 adults with elevated cholesterol found increased circulating 25-hydroxyvitamin D after supplementation with L. reuteri NCIMB 30242. However, the study did not demonstrate that the bacteria synthesized vitamin D or directly prove increased intestinal absorption. Source: Jones et al., 2013.
The evidence is also strain-dependent: another trial found no additional vitamin D benefit from the tested probiotic combination compared with vitamin-D-fortified yogurt. Therefore, improved absorption would be a hypothesis to test. Source: randomized trial of probiotic- and vitamin-D-fortified yogurt.
At a high level, the intended process would be:
Available nutrients → bacterial metabolism → vitamin D production and release → intestinal absorption.
My proposal therefore depends on discovering or engineering a pathway that functions without light under intestinal conditions. Producing a precursor alone would not count as successful vitamin D production. If this pathway cannot be established, the central concept would need redesign.
I would initially manufacture the strains separately and combine them at a measured ratio. This would make their proportions easier to control. Daily administration would be the starting model; permanent colonization would not be assumed.
3. How would I use automation?
I would use an Opentrons liquid-handling robot to prepare culture comparisons, distribute samples, and perform serial dilutions. These are established capabilities of the platform. Source: Opentrons documentation.
The conditions could compare candidate strains, community ratios, and nutrient availability. Appropriate incubation equipment would maintain the required growth environment; the liquid handler would automate sample preparation.
Separate analytical instruments would measure bacterial viability and chemically identify vitamin D. Measuring growth alone would not demonstrate production. Subsequent intestinal models would test whether the product is released and available for uptake, comparing the community with each strain alone and with a matched amount of conventional vitamin D.
At manufacturing scale, automation would shift toward bioreactor monitoring and control, harvesting, drying, blending, and capsule filling. The desktop robot would mainly support development and quality-control workflows.
4. How much would I need to make, and how long would it take?
A. A meaningful initial scale
I would define a meaningful initial scale as a 30-day supply for 1,000 people.
The intended vitamin D output would be:
1,000 people × 30 days × 5 micrograms/person/day = 150,000 micrograms = 150 mg.
That is 150 mg produced collectively inside the gut over one month, assuming the system works. The factory would manufacture the bacteria.
For the bacterial product, I would assume one capsule per day containing 10¹⁰ colony-forming units (CFU), divided equally between the two strains. CFU estimates viable bacteria capable of forming colonies. This quantity and ratio are design assumptions, not validated effective doses.
Quantity
Calculation
Result
Capsules needed
1,000 people × 30 days × 1 capsule/person/day
30,000 capsules
Viable bacteria needed
30,000 capsules × 10¹⁰ CFU/capsule
3 × 10¹⁴ CFU
Assumed culture density
10⁹ CFU/mL × 1,000 mL/L
10¹² CFU/L
Assumed overall viable recovery after processing and storage
50%
0.50
Total culture volume
(3 × 10¹⁴ CFU) / (10¹² CFU/L × 0.50)
600 L
The requirement would therefore be approximately 300 L of culture per strain. Both culture density and recovery would need experimental validation. If achievable density were ten times lower, the required volume would increase to 6,000 L.
B. Manufacturing time
For timing, assume two bioreactors, each with a 100 L working volume, operated in parallel with one strain per reactor. Three production cycles would provide 600 L total.
Manufacturing stage
Assumed time
Initial seed-culture preparation
1 day
Three cycles, each allowing 2 days for growth and 1 day for harvesting and turnaround
9 days
Final drying, blending, and packaging
2 days
Final quality-control testing
3–7 days
Total
15–19 days
These are illustrative scheduling assumptions for an established process. They exclude the potentially much longer research period needed to develop and validate the strains.
C. Required output per active producer
The production target also sets a requirement for each active producer. Suppose 10⁹ producer cells remain metabolically active for a full day:
Each would need to release approximately 5 femtograms of vitamin D per day. This calculation defines a performance target; it does not establish that the biology can achieve it.
5. What are the three biggest unknowns?
Can the bacteria actually make vitamin D?
The largest uncertainty is a complete production pathway that functions without light under intestinal conditions. I would need chemical evidence of newly synthesized vitamin D, distinguishing it from precursors or vitamin D already present in the medium.
Will production occur where absorption is possible?
Vitamin D is absorbed in the small intestine. Production farther downstream cannot automatically be assumed to provide the same benefit. I would need to measure where the bacteria are active, whether they release vitamin D, and whether the partner strain improves uptake. Source: NIH vitamin D fact sheet.
Can the community deliver a predictable, controlled output?
Diet, microbiome composition, intestinal transit, and strain survival could change production substantially. I would need to establish dose consistency, shelf stability, strain compatibility, and a way to limit activity or persistence. A small average output would not by itself establish reliable control.
AI-use disclosure
AI client: OpenAI Codex. Date: September 22, 2026.
AI was used to develop the proposal, locate scientific sources, check engineering calculations, and format the homework for this website.
Prompts:
Help me make this homework. Let’s expand on the making a bacterial comunity for the probiotic supplementation wich produces a small amount of vitamin d and vitamin d absorbtion
Ambitious in-gut production
A follow-up request asked Codex to use computer control to add the homework to this page and correct its formatting.
