1 Introduction

This tutorial demonstrates best practices for processing whole genome sequence (WGS) and mRNA sequencing (mRNA-seq) reads for allele-specific transcriptomics analysis.

1.2 Data

Sequencing data for this tutorial are described in Bresnahan et al., 2024, “Intragenomic conflict underlies extreme phenotypic plasticity in queen-worker caste determination in honey bees (Apis mellifera)”, bioRxiv. This study used instrumental insemination to create reciprocal crosses between F1 honey bees derived from different F0 genetic stocks (referred to as “lineage A” and “lineage B”) in two distinct genetic Blocks, and manipulated the F2 larvae resulting from these crosses to induce development of either the worker or queen caste fate. Whole-genome sequencing (WGS) of genomic DNA isolated from the F1 males and females used to make these crosses, and mRNA-seq of RNA isolated from the F2 larvae resulting from these crosses, was performed for quantifying and comparing allele-specific transcription between worker and queen-destined larvae.

In this tutorial, sequencing data from one genetic Block of these crosses (n=4 150x150bp paired-end WGS libraries and n=20 50x50bp paired-end stranded mRNA-seq libraries) will be used in order to reproduce Figure 4 from the study.

Figure 4. Queen-destined larvae show enriched paternal allele-biased transcription relative to worker-destined larvae. Allele-specific transcriptomes were assessed in F2 worker-destined larvae (WL) and queen-destined larvae (QL) collected from a reciprocal cross between different stocks of European honey bees. The x-axis represents, for each transcript, the proportion of lineage A reads in larvae with a lineage B mother and lineage A father (p1). The y-axis represents, for each transcript, the proportion of lineage A reads in larvae with a lineage A mother and lineage B father (p2). Each color represents a transcript which is significantly biased at all tested SNP positions: black is maternal (mat), green is lineage A, gold is lineage B, blue is paternal (pat), and grey is not significant. Center table: the number of transcripts showing each category of allelic bias and p-values for Chi-squared tests of independence for comparisons between the castes are indicated (NS = not significant). Significance of allele-biased transcription was determined using the overlap between two statistical tests: a general linear mixed model (GLIMMIX), and a Storer-Kim binomial exact test along with thresholds of p1<0.4 and p2>0.6 for maternal bias, p1>0.6 and p2<0.4 for paternal bias, p1<0.4 and p2<0.4 for lineage B bias and p1>0.6 and p2>0.6 for lineage A bias.
Figure 4. Queen-destined larvae show enriched paternal allele-biased transcription relative to worker-destined larvae. Allele-specific transcriptomes were assessed in F2 worker-destined larvae (WL) and queen-destined larvae (QL) collected from a reciprocal cross between different stocks of European honey bees. The x-axis represents, for each transcript, the proportion of lineage A reads in larvae with a lineage B mother and lineage A father (p1). The y-axis represents, for each transcript, the proportion of lineage A reads in larvae with a lineage A mother and lineage B father (p2). Each color represents a transcript which is significantly biased at all tested SNP positions: black is maternal (mat), green is lineage A, gold is lineage B, blue is paternal (pat), and grey is not significant. Center table: the number of transcripts showing each category of allelic bias and p-values for Chi-squared tests of independence for comparisons between the castes are indicated (NS = not significant). Significance of allele-biased transcription was determined using the overlap between two statistical tests: a general linear mixed model (GLIMMIX), and a Storer-Kim binomial exact test along with thresholds of p1<0.4 and p2>0.6 for maternal bias, p1>0.6 and p2<0.4 for paternal bias, p1<0.4 and p2<0.4 for lineage B bias and p1>0.6 and p2>0.6 for lineage A bias.