Quickstart guide

Getting Started

Welcome to K Pro, your AI co-pilot for biomedical research. This guide will help you understand what K Pro is and how to get the most out of its capabilities.

How to start?

1. Start with a research question in your area of interest

2. Ask about available datasets for your specific research domain

3. Request example analyses to understand what's possible

4. Try different query formulations to see how they affect results

What can you do with K Pro?

1. Visualize multiomics data with over 17 different data chart types and iterate to explore the data (e.g. with differential expression analysis)

1. Generate research hypotheses based on comprehensive biomedical data analysis

1. Access and analyze vast multimodal datasets including genomic, imaging, clinical, and literature sources using natural language queries.

2. Discover biomarkers and explore biological mechanisms

3. Optimize patient populations for clinical trials

4. Analyze competitive landscapes in your research area

5. Evaluate and enrich clinical trial data

6. Produce publication-ready visualizations and comprehensive reports

Available Data Resources

Within the K Pro interface you can explore datasets which you have access to via the Explore data tab.

K Pro free users have access to:

• TCGA Dataset: Comprehensive cancer genomics data covering over 20,000 primary cancer samples across 33 cancer types

• MOSAIC Dataset: A landmark spatial omics initiative providing unprecedented tumor structure information across multiple cancer types (K Pro account subscribers)

Key Agents Working Together

K Pro integrates several specialized agents that work together seamlessly:

1. Knowledge Agent: Provides access to biomedical knowledge bases and literature

2. MultiOmics Agent: Analyzes complex biological datasets across different modalities

3. Literature Review Agent: Searches and synthesizes relevant scientific publications

4. Visualization Agent: Creates clear, publication-ready visualizations of your data

5. Clinical Data Agent: Correlates findings with patient characteristics

Exploring Capabilities

1. Start with a research question in your area of interest

2. Ask about available datasets for your specific research domain

3. Request example analyses to understand what's possible

4. Try different query formulations to see how they affect results

Prompting Best Practices for Optimal Results

1. Be specific in your queries: The more precise your question, the more targeted the response

2. Start broad, then narrow down: Begin with exploratory questions before diving into specifics

3. Specify data types of interest: Mention whether you're interested in genomic, clinical, or other data types

4. Request visualizations: Ask for graphs or plots to better understand complex relationships

5. Follow up on insights: Use the suggested follow-up questions to explore topics further

What to Avoid

1. Overly vague queries: "Tell me about cancer" is less effective than "Show me survival differences between EGFR-mutated and wild-type lung adenocarcinoma patients"

2. Requesting unsupported analyses: If a specific analysis isn't available, K Pro will suggest alternatives

3. Ignoring citation links: The platform provides source links for verification and further reading

Note: K Pro is designed to be your collaborative research partner, helping you navigate complex biological data and accelerate discoveries in biomedical research. As you become more familiar with the platform, you'll discover increasingly powerful ways to leverage its capabilities for your specific research needs.