Tutorials

Explore GraphFLA with nine datasets, from chemical reactions and materials to neural architectures and compiler settings. Each tutorial takes you from preparing the data to building a landscape and interpreting its analysis.

Choose an optimization problem below. Every page includes Python code, saved results, a downloadable notebook with its data, and a link to run it in Google Colab.

Tutorial	Study / dataset	Variables	Objective	Colab
Chemical reaction optimization	Suzuki–Miyaura coupling	Ligand, base and solvent	Maximise UV product-area percentage
Reaction process optimization	Electrochemical flow hydrogenation	Concentration, temperature and residence time	Maximise the reported integrated index
Alloy composition optimization	Tungsten–rhenium–osmium alloys	Tungsten and rhenium fractions, with osmium as the remainder	Maximise high-temperature hardness
Materials design	Hybrid perovskites · ABX₃	Organic ion, metal and halide	Minimise calculated electronic band gap
Microbial growth optimization	BacPUS · Gut bacteria and polysaccharides	Bacterial strain and polysaccharide	Maximise mean growth at 48 hours
Enzyme inhibitor discovery	Cyanimide library · Mouse USP18	Amine and carboxylic-acid building blocks	Maximise mUSP18 inhibition
Drug combination optimization	NCI-ALMANAC · Cancer cell assays	Partner drug and two dose indices	Minimise percent growth in a fixed cell-line context
Neural architecture search	NAS-Bench-201 · CIFAR-10	Operations on six architecture edges	Maximise validation accuracy for a fixed training seed
Software configuration tuning	LLVM · Compiler flags	Ten Boolean compiler options	Minimise measured compilation time