The paper discusses the problem of clustering multi-condition gene expression patterns, aiming to understand gene function and regulatory mechanisms. It introduces a stochastic model for input data and presents an efficient algorithm that recovers cluster structures with high probability in $O(n \log(n))^c$ time, where $n$ is the number of genes. The algorithm is based on finding the nearest clique graph to the input graph, where edges are randomly removed or added. The paper also proposes a practical heuristic approach, the Cluster Affinity Search Technique (CAST), which uses average similarity to cluster genes. The performance of both the theoretical and practical algorithms is evaluated through simulations and applied to real gene expression data, demonstrating their effectiveness in recovering underlying cluster structures.The paper discusses the problem of clustering multi-condition gene expression patterns, aiming to understand gene function and regulatory mechanisms. It introduces a stochastic model for input data and presents an efficient algorithm that recovers cluster structures with high probability in $O(n \log(n))^c$ time, where $n$ is the number of genes. The algorithm is based on finding the nearest clique graph to the input graph, where edges are randomly removed or added. The paper also proposes a practical heuristic approach, the Cluster Affinity Search Technique (CAST), which uses average similarity to cluster genes. The performance of both the theoretical and practical algorithms is evaluated through simulations and applied to real gene expression data, demonstrating their effectiveness in recovering underlying cluster structures.