2.4. Expression profiling based on long-reads
The long-read digital MinION signals were converted from POD5 to the
FAST5 format using the pod5-file-format program
(https://github.com/nanoporetech/pod5-file-format).
Next, the transcriptomic sequences were basecalled by Guppy v.6.0.0
(https://nanoporetech.com/support). The FASTQ raw reads were
quality-checked and passed to the mapping steps (as a referenceRiccia fluitans genome), supported by minimap2 v.2.26 software
with default parameters . Similar to short-reads analysis, the gene
expression profiles produced by the long-reads sequencing method were
also estimated using stringtie, featureCounts and DESeq2 softwares. For
transcript level expression quantification, the above proceed BAM files
were used again by bambu v3.2.4 software to estimate the transcript
count expression matrix for multiple samples . The differentially
expressed genes (DEGs) and differentially expressed transcripts (DETs)
statistical significance was determined with the following parameters:
padj < 0.05 and absolute log2FoldChange > 1. The
results from both methods (short - and long-reads) were intersected and
only common results were considered as final transcriptomic DEGs and
DETs results. Additionally, the transcriptomic sequences were divided
into coding and non-coding groups. Two potential coding prediction
softwares, CPC2 v.1.0.1 and PLEK v.1.2 , classified transcripts into
separate groups. According to those classifications, significant genes
were named differentially protein-coding genes and differentially long
non-coding RNAs (DELs). If there were discrepancies in identification of
coding potential between the two programs, those RNA were signed as
OtherRNA. Relationships between DEGs, DELs and OtherRNA were estimated
by co-expression analysis. Pairs of DEGs-DELs, DEGs-OtherRNA, and
DELs-OtherRNA with similar transcriptomic profiles were characterized
based on the Pearson correlation coefficient (r > 0.8 and p
< 0.05). The results were visualised using the ggplot2 v.3.4.4
and circlize v.0.4.15 R Bioconductor v.3.18 packages.