I started using Obsidian two years ago to organize research notes scattered across notebooks, Word documents, and loose PDFs. Six months in, I realized I’d rebuilt a searchable, interconnected repository of my research thinking that I actually used. The key was getting the structure right from the beginning.
Obsidian is a note-taking application built on local Markdown files. Unlike cloud-based tools, your notes live on your computer, sync across devices however you want (or not at all), and never disappear if a company shuts down. For scientists, this matters. You need notes that outlive the tools you’re currently using.
This guide walks you through setting up Obsidian specifically for research: papers you’ve read, experiments you’ve run, projects you’re planning, and ideas that connect across them. This isn’t a philosophical essay on knowledge management. It’s a setup guide with specific folder structures, templates, and configurations that work for scientists.
Section 1: Why Obsidian for Research
Before diving into setup, let’s clarify why Obsidian is worth your time.
Longevity. Your notes are plain Markdown files on your hard drive. Markdown has existed for 20 years and isn’t going anywhere. Obsidian is the application layer, but your actual notes aren’t locked into it. You can read them in any text editor.
Search. Obsidian indexes all your notes and lets you search across them instantly. If you note that you used a particular imaging protocol in an experiment from two years ago, you can find that note in seconds.
Backlinks. This is Obsidian’s biggest differentiator. When you link notes together (e.g., linking a paper note to a project note), Obsidian tracks those connections. This creates a graph of your research landscape. You can see which papers relate to which projects, which experiments connect to which findings.
No subscription, no lock-in. Obsidian’s core is free. Paid features (sync, publish) are optional. You own your data.
Privacy. Your notes never leave your computer unless you choose to sync them. No cloud backend reading your research notes.
For a wet lab scientist, this means your experiment notebook lives in Obsidian, searchable and versioned. For a computational biologist, your pipeline notes, analysis decisions, and literature notes are all connected and indexed. For anyone writing a thesis or grant, your literature review becomes a knowledge base you can search and reorganize.
Section 2: Installation and Basic Setup
Download and install Obsidian:
Go to obsidian.md and download the version for your operating system (macOS, Windows, or Linux). Install it like any other application.
Create your first vault:
When you open Obsidian, it asks you to open or create a vault. A vault is a folder on your computer where all your notes live. Create a new vault. Call it something like “Research” or “Science Notes”. Choose a location you’ll remember and where you have backup systems in place (cloud backup, external drive, or Git).
Tour the interface:
Obsidian’s main interface has:
- A left sidebar showing your folder structure and any bookmarked searches.
- A center panel where you write and read notes.
- A right sidebar showing backlinks (other notes linking to this one), related notes, and other panels.
Open Obsidian’s built-in tutorial (“Open” button, then “Open sample vault”) to familiarize yourself with creating notes, linking them, and using the search. Spend 15 minutes here. It’s worth it.
Enable core plugins:
In Settings (bottom left), go to “Core Plugins”. Make sure these are enabled (they usually are by default):
- Backlinks (shows links to this note)
- Graph view (visualizes your note network)
- Daily Notes (creates a note for today)
- Templates (allows template snippets)
- Starred (bookmark frequently accessed notes)
Section 3: Folder Structure for Scientists
Your folder structure is the skeleton of your system. Here’s a structure that works for most research contexts:
Research/
├── Papers/
├── Experiments/
├── Projects/
├── Daily/
├── Ideas/
└── Templates/
Papers/ stores one note per paper you’ve read. Each note contains the abstract, key findings, methodology, and your thoughts.
Experiments/ stores notes from your lab work. If you’re a wet lab scientist, this is your experiment notebook (raw data, observations, follow-up questions). If you’re computational, it’s your analysis runs and pipeline execution logs.
Projects/ stores project-level planning. One note per active project. Links to the relevant papers and experiments for that project.
Daily/ contains a daily note (one per day). This is where you capture the day’s todos, quick observations, and thoughts. Daily notes often reference papers and experiments, creating automatic backlinks.
Ideas/ captures half-baked ideas, research directions you’re considering, methods you want to try. No structure required yet; you’ll organize and connect them later.
Templates/ stores Obsidian templates (not your actual notes, but template files you’ll use to create new notes).
You can expand this structure. Add a “Courses/” folder if you’re learning new skills. Add “Talks/” if you prepare seminars. Add “Reading/” if you maintain reading lists. The key is consistency: each note type lives in one place.
Section 4: Literature Note Template
When you read a paper, create a note in Papers/. Use this template consistently. Consistency means you can search and query across all your paper notes later.
Create a file called Paper Template.md in Templates/ with this structure:
# {{title}}
**Authors:**
**Year:**
**DOI:**
**Journal:**
**URL:**
## Key Finding
[One sentence summary of the main result.]
## Methodology
[What did they do? What model organism, cell type, or computational method?]
## Key Results
- [Result 1]
- [Result 2]
- [Result 3]
## Limitations
[What are the weaknesses or limitations the authors acknowledge or that you notice?]
## Personal Reaction
[What did you think? How does this relate to your work? Any questions?]
## Connections
[Links to other papers, projects, or ideas. Format: [[paper name]] or [[project name]]]
## Tags
#literature #methodology #cancer-immunotherapy
How to use this template:
In Obsidian, when you create a new note, use the “Templates” button (or command palette: Ctrl+P, search “Templates: Insert template”) and select “Paper Template”. The template populates with placeholders.
Fill in the fields as you read. The “Key Finding” should be one sentence. The “Key Results” should be bullet points, not paragraphs. The “Personal Reaction” is where you think critically: Does this contradict something you read elsewhere? How does it change your understanding?
The “Connections” section is crucial. Link to other papers using [[paper name]] syntax. Link to projects using [[project name]]. These links create the backlink graph that makes Obsidian powerful.
Concrete example:
# Circulating Tumor DNA as a Biomarker for Colorectal Cancer Recurrence
**Authors:** Smith et al.
**Year:** 2023
**DOI:** 10.1038/s41591-023-01234-z
**Journal:** Nature Medicine
**URL:** https://doi.org/10.1038/s41591-023-01234-z
## Key Finding
Early detection of ctDNA in post-treatment blood predicts colorectal cancer recurrence before standard imaging can detect it.
## Methodology
Prospective cohort of 150 colorectal cancer patients post-resection. ctDNA quantified using deep sequencing of patient-specific mutations. Follow-up with imaging and clinical assessment over 24 months.
## Key Results
- Patients with detectable ctDNA had 5-fold higher recurrence risk at 6 months
- ctDNA detection preceded imaging detection by average 3.2 months
- Specificity 96%, sensitivity 87% for recurrence prediction
## Limitations
- Single institution, relatively small cohort
- Only one sequencing platform tested
- No comparison to CEA or other standard biomarkers
## Personal Reaction
This is directly relevant to my current project on using ctDNA for early detection. The 3-month lead time over imaging is significant. I should check if they validated their mutation selection strategy.
## Connections
[[ctDNA and cancer monitoring]]
[[colorectal cancer detection project]]
[[sequencing depth and sensitivity]]
## Tags
#literature #ctdna #colorectal-cancer #biomarker
Section 5: Experiment Note Template
For your experiments, use a similar template. Create Experiment Template.md:
# Experiment: {{title}}
**Date:**
**Aim:**
**Hypothesis:**
## Protocol
[Link to protocol document or brief description of method used. [[protocol name]] if you store protocols in Obsidian.]
## Materials
- [Material 1]
- [Material 2]
## Procedure
[Step-by-step execution, any deviations from protocol.]
## Results
[Raw data location, summary of key measurements, observations.]
## Observations
[Anything unexpected? Any issues?]
## Interpretation
[What do the results mean? Do they support the hypothesis?]
## Follow-up Questions
[What should you test next?]
## Related Papers
[[paper 1]]
[[paper 2]]
## Tags
#experiments #wet-lab #imaging
Concrete example:
# Experiment: CAR-T Cell Expansion with IL-2 Supplementation
**Date:** 2025-12-01
**Aim:** Test whether daily IL-2 supplementation increases CAR-T cell expansion vs. standard protocol
**Hypothesis:** IL-2 supplementation will increase cell viability and expansion at day 10
## Protocol
[[CAR-T expansion protocol v3]]
## Materials
- CAR-T cells (patient sample A123, passage 2)
- IL-2 stock (Peprotech, 200 IU/ul)
- X-VIVO media (Lonza)
- 6-well plates
## Procedure
Day 0: Plate CAR-T cells at 1x10^6 cells/ml.
Treatment group: Add 1000 IU/ml IL-2 daily (diluted in PBS, added to wells).
Control group: PBS addition only, same volume.
Incubate at 37C, 5% CO2.
Day 3, 6, 10: Count viable cells, measure viability via trypan blue.
## Results
Day 10 viable cell counts:
- Control: 8.2 x 10^6 cells (viability 87%)
- IL-2 treatment: 14.5 x 10^6 cells (viability 92%)
Fold expansion: Control 8.2x, IL-2 14.5x
## Observations
IL-2 group showed slightly higher metabolic activity by MTS assay on day 7. No apparent morphology changes. One control well contaminated on day 5, excluded from analysis.
## Interpretation
IL-2 supplementation increased expansion approximately 1.8-fold, supporting the hypothesis. Viability also slightly higher in treatment group. Next: repeat with 3 patient samples to confirm reproducibility.
## Follow-up Questions
- Does higher concentration of IL-2 improve further?
- What is the optimal frequency of IL-2 addition?
- How does this scale to bioreactor conditions?
## Related Papers
[[IL-2 signaling in T cell expansion]]
[[Cytokine supplementation CAR-T manufacturing]]
## Tags
#experiments #CAR-T #cell-culture
Section 6: Useful Plugins
Obsidian’s power multiplies with plugins. Core plugins are built-in and free. Community plugins are created by users and extend Obsidian’s functionality.
Enable these core plugins:
- Backlinks: Shows other notes linking to this note.
- Graph view: Visualizes your note network as a graph.
- Daily notes: Creates a note for today automatically.
- Templates: Allows inserting templates into notes.
- Starred: Bookmark frequently accessed notes.
Install these community plugins:
Go to Settings > Community Plugins > “Browse” to find and install these:
Dataview: This plugin lets you query your notes like a database. For example, you could create a page that automatically lists all papers tagged #cancer-immunotherapy, or all experiments from the past month. For scientists, this is incredibly useful. Once installed, you write queries in code blocks:
LIST
FROM #literature
WHERE contains(DOI, "2023") OR contains(DOI, "2024")
SORT file.name DESC
This lists all literature notes with DOIs from 2023 or 2024, sorted by filename (newest first). You can build queries for “papers I’ve read in the past month”, “experiments with high viability”, etc.
Templater: More powerful than the built-in Templates plugin. Allows dynamic fields like date, file name, and conditional logic. Example: insert today’s date automatically with <% tp.date.now("YYYY-MM-DD") %>.
Zotero Integration: If you use Zotero (a free reference manager), this plugin imports your Zotero library into Obsidian. You can insert citations and links to papers directly from Zotero. This is a game-changer for researchers.
Citations: Allows inserting citations in various formats (APA, MLA, Chicago) directly from your literature notes.
Obsidian Web Clipper: A browser extension that clips articles and web content into Obsidian. When you find a paper abstract or method description online, clip it directly into your vault.
Tasks: If you want to track todos and project management in Obsidian, this plugin adds task list functionality.
Start with Dataview and Templater. These two give you the most immediate value for research notes. Install others as your workflow matures.
Section 7: Sync and Backup
Your notes must be backed up. Obsidian doesn’t backup automatically unless you use their paid Obsidian Sync service. Here are your options:
Obsidian Sync (paid): Cloudflare-powered sync across your devices. Costs about $10/month. Your notes are encrypted end-to-end. Simple to set up (enable in Settings > Sync), works seamlessly. If you have multiple devices and want hassle-free syncing, this is the easiest option.
iCloud or Dropbox (free): Place your vault folder in iCloud Drive or Dropbox. Both services sync folders across devices. This works well, though occasional sync conflicts can happen if you edit the same note on two devices simultaneously. Dropbox is more reliable for this use case than iCloud.
Git (free, but requires comfort with Git): If you’re familiar with Git, initialize your vault as a Git repository and push to GitHub (private repository). This creates version history, so you can roll back to any previous version of a note. It also syncs across devices if you pull/push regularly. The community plugin “Obsidian Git” automates this (commits and pushes automatically).
Local backup: At minimum, set up automated backup of your vault folder to an external drive. Use Time Machine (macOS) or Backblaze/similar (Windows/Linux).
My recommendation: If you’re comfortable with Git, use Git plus a private GitHub repository. You get version history and free syncing. If you prefer a simpler setup, use Dropbox or Obsidian Sync. At minimum, never rely solely on local storage without backup. Research notes are too valuable to lose.
Section 8: Building Your First Note
Now that everything is set up, create your first literature note. Pick a paper you’re currently reading. Create a new note in Papers/, use the Paper Template, and fill it out as you read.
As you read, look for connections to other papers you’ve read or projects you’re working on. Use [[double brackets]] to link them. Don’t worry about perfect linking yet; rough connections are fine. Over time, your notes become more connected and the backlink graph becomes more useful.
After three weeks of capturing papers and experiments, open the Graph View (button in top right, or command palette). You’ll see your notes connected in a visual network. This graph is your research landscape. It shows you which topics cluster together, which papers you’ve read about a particular experiment, which projects relate to which literature.
After a month, try creating a Dataview query. Go to your Projects folder and create a note called “Literature Summary”. Use a Dataview query to auto-populate a list of papers tagged with your project:
LIST
FROM #my-project
SORT file.name DESC
This automatically lists all notes tagged with #my-project in the order you read them. You can embed this in your project notes. Now your project notes automatically pull in relevant literature and experiments.
Next Steps
Once your Obsidian vault is set up and you’ve created 5-10 literature notes, the system starts paying dividends. The next step is to create your first project note. Link it to relevant papers and experiments. Watch how the backlink graph shows you the landscape of your research.
From there, build toward a weekly review practice. Every Friday, spend 15 minutes reviewing your weekly experiment notes and tagging them. Over months, your vault becomes a searchable, interconnected knowledge base that’s more valuable than any folder of PDFs.