2.5 | Data analysis
To determine how the libraries needed to be normalized and pooled for HiSeq sequencing we used initial read counts of MiSeq data mapped to fungal and cyanobacterial reference genes (β-tubulin (AFJ45056.1) and glycerol 3-phosphate dehydrogenase (AFJ45057.1) for fungus; protein translocase subunit secA (CP026681.1; region: 4141894 - 4142180) and RNase P RNA gene rnpB (CP001037.1; region: 1485004 - 1485242) for cyanobacteria). The quality of the MiSeq and HiSeq data was assessed using FastQC version 0.11.5 (https://www.bioinformatics.babraham.ac.uk/projects/fastqc/; accessed: 15.12.2019) and MultiQC version 1.1 (https://multiqc.info/; accessed: 25.02.2019). Poor quality base reads were removed with the FASTX-toolkit version 0.0.13 (http://hannonlab.cshl.edu/fastx_toolkit/; accessed: 18.02.2019). Adapter sequences were trimmed with Trimmomatic version 0.36 (http://www.usadellab.org/cms/?page=trimmomatic; accessed: 20.02.2019). The processed paired-end MiSeq data was used for de novo transcriptome assembly with Trinity software version 2.4.0 (Haas et al., 2013). The quality of the assembly was assessed with the Trinity perl script TrinityStats.pl. The HiSeq data was pseudoaligned to thede novo transcriptome assembly with the RNA-seq quantification program kallisto version 0.45.0 (Bray et al., 2016). Coding regions were identified with TransDecoder (http://transdecoder.github.io; accessed: 20.03.2019). For the respective parameter settings see Electronic Supplementary Table S1.