v0.0.1-alpha

First pre-release candidate!

• Estimates "denoised" log-fold-change effects of each perturbation each gene from sparse single cell count data
• Prioritizes high-confidence findings using Bayesian uncertainty estimates
• Borrows information across guides targeting the same genetic locus while simultaneously identifying high/low-effect guides
• Uses the familiar scvi-tools interface to empower users to quickly analyze their data in standard formats (AnnData/MuData)