Polyketide production remains at a very low level (~10 mg/l) in Escherichia coli,partly owing to the discordance between foreign biosynthetic modules and native metabolic modules.However, it remains a challenge to characterize the module interaction of native and heterologous metabolism.Here, an omics approach based on in silico genome-scale metabolic analysis and comparative transcriptome was developed to systematically investigate the problematic interactions between erythromycin precursor 6-deoxyerythronolide B (6dEB) module and native modules in E.coli.In comparison with in silico 'ideal' situation for maximum 6dEB biosynthesis rate, the 'actual' situation represented by transcriptome profiling showed tremendous differences in the flux distribution of metabolic modules, especially at central metabolic modules and competitive macromolecule biosynthesis modules.Potential targets were identified for improving 6dEB biosynthesis based on this omics analysis.All 25 predicted targets at modules of pentose phosphate (PP) pathway and nucleic acid biosynthesis were first tested for 6dEB production in E.coli;engineered 18 targets led to >20% increase in 6dEB yield, especially that knockdown of talA,talB, cysQ, and guaB enhanced 6dEB titer by 92%, 135%, 91%, and 125%,respectively.Co-repression of cysQ and guaB led to a 183.1% increase in 6dEB production (150.4 mg/1 of 6dEB in flask) compared to the control pACYCDuent-1.Our results suggest that improving the interactions and adaptations among the modules to meet with those of the ideal situations at systemic levels would be an effective strategy to improve the yield of heterologous products.