add calculation of transcript abundance assay and protocol

assays/CalculationOfTranscriptAbundance/protocols/CalculationOfTranscriptAbundanceUsingKallisto.md

+## Calculation of transcript abundance using Kallisto
+
+The program Kallisto (v.0.46.0) was used to estimate the relative expression of genes in tomato and potato (Bray et al., 2016). As a first step, the raw sequence reads were compared to the transcript sequences. This step in Kallisto is designated as the pseudoalignment step. To improve the quality of the pseudoalignment, low-quality reads and adapters were removed from the transcriptomes using Trimmomatic under the following settings: seed mismatch = 2; palindrome clip threshold = 30; simple clip threshold = 10; LEADING = 3; TRAILING = 3; SLIDINGWINDOW= 4:15; MINLEN =36 (Bolger et al., 2014; Figure S2). Subsequent quality controls were performed using FastQC (Andrews, 2010). As Kallisto requires information on fragment length for single-end sequenced transcriptomes, the fragment length denoted by the authors was used. If this information was not available, the recommended fragment length of the reported RNA isolation kit was used. The standard deviation was set to ±17.5 bp. Kallisto indices (used for generating the pseudoalignments) were based on the tomato ITAG4.0 and the potato PGSC_DM_v4.03 genome releases. *R*-genes missing from the current genome releases were manually added to the list of transcripts (indices in Kallisto). Transcript abundance was calculated as transcripts per million (TPM; Wagner et al., 2012). We chose to use TPM since it normalizes the transcript abundance for gene length and library size, making TPM values comparable across experiments. Genes for which the TPM values were less than 1 were treated as “off” and for these genes, TPM was set to zero. All scripts and settings used for these analyses are available at the following site: https://github.com/LauraERose/LargeScaleTranscriptomeAnalysis.
\ No newline at end of file