Monthly Archives: August 2013

Six new RNAseq datasets from Chalara fraxinea cultures and detection of mitoviruses infecting Chalara fraxinea


Rachel Glover, Ian Adams, Paul Beales. Food and Environment Research Agency, York, UK.



Mixed material RNASeq assemblies

AT1, AT2, ATU, Upton Broad, Holt Park

Chalara fraxinea mycelia RNASeq assemblies

Kenninghall Wood, Fera 88, Fera 93, Fera 94, Fera 105, Fera 232, Fera 233


The AT1, AT2, ATU, Upton and Holt mixed-material RNASeq assemblies were searched against the Genbank nr database with blastx in order to investigate the virome of each sample. A number of novel viral sequences were detected. A blastx was then carried out against the Kenninghall wood (KW) mycelia assembly to determine whether any of these viral sequences were likely to be infecting Chalara fraxinea. While none of the novel viruses originally discovered in the other datasets were present in the KW assembly, a 2.4kb mitovirus was detected. A tblastx search of this mitovirus sequence against the mixed-material samples discovered shorter contigs in the Holt and Upton assemblies which had homology to the KW mitovirus but which had not been detected in the original blastx against Genbank.
Fera has been testing large numbers of samples as part of the response to the presence of Ash Dieback in the UK and so we sequenced six cultured isolates of Chalara fraxinea (RNASeq with MiSeq) to investigate the prevalence of mitoviruses in a small number of our cultures. Mitoviruses were detected in the Fera 88, 93 and 94 samples but not Fera 105, 232 and 233.



Figure: Mitovirus positive samples are shown with red pins, negative with white pins.


The presence or absence of mitovirus does not appear to be geographically correlated but the Fera samples were not comprehensively sequenced so mitoviruses may be present in these samples but not sequenced. Further work is required to investigate whether these mitoviruses confer hypovirulence to Chalara fraxinea.

Fraxinus Game

Play Fraxinus

A new Facebook game that could help save the Ash tree.
Players help analyse real genetic data from the fungus which causes Chalara ash dieback and from common the ash, Fraxinus excelsior.
By matching and rearranging patterns, each player will contribute a small bus useful analysis to the bid to save the Ash tree.

mRNA-Seq analysis of Tree 35

The contributors

Martin Trick (JIC), Andrea Harper (JIC*), Leah Clissold (TGAC) and Ian Bancroft (JIC*)

*Present affiliation: University of York

The material

First flush leaf material was harvested from Tree 35 in Denmark in May 2013.

The analysis

mRNA was extracted and then a strand-specific, paired-end Illumina RNA-Seq library constructed. Around 193 million read pairs were obtained from a single HiSeq 2500 lane and these are now available from The Sainsbury Laboratory’s FTP server, with details in the github repository here.

Trinity was used to assemble transcript fragments from a down-sampled set of 100 million pairs of reads, generating an unusually large number of 517,056 assemblies (data/tree/master/ash_dieback/fraxinus_excelsior/tree35/assemblies/mRNA/tree35_trinity.fasta.gz. RSEM transcript abundance analysis was then carried out to select the 238,283 principal isoforms from these (data/tree/master/ash_dieback/fraxinus_excelsior/tree35/assemblies/mRNA/tree35_principal_isoforms.fasta.gz), and this will serve as our reference sequence for the association work that will soon be conducted on a diverse panel of trees from Denmark.

Candidate open reading frames were extracted and the predicted peptides (data/tree/master/ash_dieback/fraxinus_excelsior/tree35/annotations/mRNA/tree35_predicted_peptides.fasta.gz) were queried with BLASTP against the protein database, creating a first draft functional annotation (data/tree/master/ash_dieback/fraxinus_excelsior/tree35/annotations/mRNA/annotated_peptides.txt). A significant number of proteins apparently originating from Bradyrhizobium sp. BTAi1 were found – the validity or significance of this is currently unknown. Because of this issue and the unusual complexity of the assemblies, these data should be treated with caution; another Tree 35 sample is due to be sequenced soon. Finally, the transcript assemblies were queried with BLASTN against the Tree35 genome scaffolds developed by TGAC. Around 75% of the transcripts were unambiguously located to scaffolds, thus creating an extra layer of annotation for the genomic assembly (data/tree/master/ash_dieback/fraxinus_excelsior/tree35/blast/mRNA/tree35_transcripts_vs_Nornex_s1v1.gz).