ELDA: Real-Time Drug Response Profiles for 25 Pediatric Leukemia Samples

The dataset describes ELDA, a drug screening platform that generated functional drug response profiles for 25 pediatric acute lymphoblastic leukemia (ALL) samples against a panel of 50 agents. It was created by Samara Sousa Mariano and involves time-lapse microscopy data analyzed to quantify cell viability over 96 hours.

Use Cases

• Train a computer vision model on time-lapse microscopy images to quantify cell viability for drug screening.
• Analyze drug sensitivity and resistance patterns across the 50-agent panel for the 25 pediatric ALL samples.
• Investigate correlations in activity profiles for drugs with similar mechanisms of action within the dataset.
• Validate image analysis algorithms for detecting subtle cell membrane movements in a 3D co-culture microenvironment.

Strengths

• Profiles 25 pediatric ALL patient samples, providing a basis for patient-specific analysis.
• Tests a panel of 50 clinically relevant and experimental drugs per sample.
• Method validated against established cytotoxicity assays (MTT, Calcein-AM) and supported by in vivo validation.

Limitations

• The core data (e.g., raw images, quantified viability time series) is embedded within a 5.9 MB PDF, requiring extraction for computational analysis.
• Sample size is limited to 25 pediatric ALL cases, which may restrict statistical power for some analyses.
• The dataset is static, representing a specific research study with no stated update frequency.

Provenance

Source

figshare, authored by Samara Sousa Mariano.

Collection Method

Data generated via the ELDA platform using time-lapse bright-field microscopy of primary ALL cells in a 3D co-culture, with custom image analysis.

Freshness

Last updated 2026-03-20.