Failure in running BWA on human fusion genome

[Gigiux]

Dear all,
I have prepared a fusion genome by concatenating the human genome available here:

and a virus genome obtained by merging several individual virus genomes together. The headers of the NCBI genome are like this:

>chr1  AC:CM000663.2  gi:568336023  LN:248956422  rl:Chromosome  M5:6aef897c3d6ff0c78aff06ac189178dd  AS:GRCh38
>chr2  AC:CM000664.2  gi:568336022  LN:242193529  rl:Chromosome  M5:f98db672eb0993dcfdabafe2a882905c  AS:GRCh38
>chr3  AC:CM000665.2  gi:568336021  LN:198295559  rl:Chromosome  M5:76635a41ea913a405ded820447d067b0  AS:GRCh38
[...]

>chrUn_GL000218v1  AC:GL000218.1  gi:224183305  LN:161147  rl:unplaced  M5:1d708b54644c26c7e01c2dad5426d38c  AS:GRCh38
>chrEBV  AC:AJ507799.2  gi:86261677  LN:171823  rl:decoy  M5:6743bd63b3ff2b5b8985d8933c53290a  SP:Human_herpesvirus_4  tp:circular

Thus I prepared the header of my sequence as:

>chrV  AC:XXXXXXXX.1  gi:00000000  LN:370064105  rl:Chromosome   M5:5aa5be7025d7baa666a8651e0909e4ce  AS:1  SP:All_viruses  tp:linear

before concatenating with the NCBI's.

I prepared the indices with

bwa index fusion.fa

and then ran the alignment with

bwa mem -t 8 -R $rd_grp $Fusion $R1 $R2 | samtools sort -o ${out}SRT.sam

where $rd is the read group header,
$Fusion is the path to the fusion.fa reference,
$R1/2 are the fastq files of interest and
${out}SRT.sam is the alignment output.

However, I got this error:

[bns_restore_core] Parse error reading /.../fusion.fa.amb

On internet I found that the error might be due to spaces in the sequence, so I applied:

sed -i 's/\s*$//g' fusion.fa
sed -i 's/^[^>]\s*$//g' fusion.fa

and also

printf "chrV \tAC:XXXXXXXX.1\tgi:00000000\tLN:370064105\trl:Chromosome   M5:5aa5be7025d7baa666a8651e0909e4ce\tAS:1\tSP:All_viruses\ttp:linear" > fusion.fa

on the idea that the fields cuold be tab delimited. The error persisted with each version.

May I ask you if you have some hint on this error and how to sort it out.
Thank you.