Concatenated Genome Strategy Outperforms Sequential Mapping for Accurate Dual RNA-Seq in Fungal Cocultures

Dual RNA-seq enables simultaneous transcriptome profiling of interacting organisms. However, accurate read assignment remains challenging in fungal cocultures due to phylogenetic proximity and shared homologous sequences. We compared efficiency and robustness of sequential and composite reference mapping strategies to optimize dual RNA-seq analysis of a fungal coculture on sugarcane bagasse. RNA-seq samples from a fungal (Pleurotus citrinopileatus and Trichoderma reesei) co-cultivated on sugarcane bagasse were sequenced and quality-filtered using FastP and FastQC. For in silico read separation, two mapping strategies were evaluated using the STAR aligner: 1) Sequential Mapping (seq): A two-step alignment where reads were mapped to primary genome A, and unmapped reads were subsequently aligned to secondary genome B (with inverted order to assess alignment bias); 2) Concatenated Mapping (conc): A single-step strategy where reads were aligned against a combined reference genome (FASTA files merged via cat command), allowing direct competition between species for read alignment. Strategy performance was evaluated using custom Bash scripts based on standardized comparison metrics, including concordant reads, cross-mapping (mismatches), multi-mapping, and rescued false negatives. Illumina NovaSeq 6000 sequencing generated 175 million high-quality reads, achieving mapping rates above 96% across all strategies. While no significant differences were observed in uniquely mapped reads regarding the strategy or genome alignment order in the sequential approach (p = 0.976 and p = 0.999, respectively), the concatenated reference genome provided a more balanced read distribution. Cross-mapping metrics was consistently low -below 2% across all samples. The number of reads classified as “Cross-mapped (seqA vs concB)” (0.67%−2.01%) exceeded those in “Cross-mapped (seqB vs concA)” (0.01%), indicating a higher frequency of mismatches or misassignments when Pleurotus served as genome A in the sequential mapping strategy. Notably, the concatenated approach substantially minimized cross-mapping and multi-mapping compared to the sequential strategy, which proved highly sensitive to species alignment order. False positives reads in the sequential approach that were converted to true positives in the concatenated approach, ranged from 26% to 71% across the mapping orders and were consistently higher in the PT (Pleurotus first) sequential order compared with the TP (Trichoderma first) sequential order. Furthermore, concatenated mapping reduced computational time by approximately 93%. Together, these results establish concatenated genome mapping as a highly efficient and accurate framework for resolving mixed fungal transcriptomes. This approach enables the quantification of host gene expression changes that underpin biotechnological advancements, including biofuel production, pharmaceuticals, and biopesticide development, thereby facilitating the identification of candidate genes to engineer elite strains of lignocellulolytic fungi.