Updated analysis: Cleaning BRAF fusions for LGAT subtyping

[kgaonkar6]

What analysis module should be updated and why?

molecular-subtyping-LGAT
In the LGAT subset we see known KIAA1549--BRAF fusion and some other BRAF fusions as you can see in FusionName column here:
https://github.com/AlexsLemonade/OpenPBTA-analysis/blob/master/analyses/molecular-subtyping-LGAT/results/lgat_cleaned_all_table.tsv

So @jaclyn-taroni and I want to discuss here if any BRAF fusions found in LGAT sample should be included in molecular_subtype== "LGG,BRAF fusion" or would the fused BRAF need to satisfy other conditions like kinase domain retention and overexpression as mentioned in for the oncogenic KIAA1549--BRAF fusion
https://www.frontiersin.org/articles/10.3389/fonc.2015.00054/full
"The constitutive kinase activity of KIAA1549:BRAF fusion oncoprotein is due to the loss of the BRAF N-terminal auto-inhibitory domain which usually regulates BRAF activity (15, 28). Lin et al. demonstrated that the KIAA1549:BRAF fusion is transcribed from the KIAA1549 gene promoter and thus is expressed at higher levels than wild-type BRAF, so the fusion oncoprotein is not only constitutively active but also over expressed giving two mechanisms of aberrant activity (27)."

A example of other BRAF fusion can be seen in this sample: has no BRAF V600E mutation but does have two fusions (BRAF--RNF130_, RNF130--BRAF):

BS_9X9E86DY	No	PT_20ZM6THA	7316-3629	BS_E94QRJDW	BRAF--RNF130, RNF130--BRAF

Here I've plotted the breakpoints for BRAF gene in each fusion to show the differences in kinase domain retention status per fusion
RNF130--BRAF is a cosmic fusion https://cancer.sanger.ac.uk/cosmic/fusion/overview?fid=199318&gid=296121. In-frame fusion and also retains the kinase domain because the breakpoint is before the kinase domain

BRAF--RNF130 is the reciprocal frame-shift fusion that doesn't fully retain the kinase domain as you can see the breakpoint is within the kinase domain

What changes need to be made? Please provide enough detail for another participant to make the update.

Should we add a notebook/script before 01-make-lgat-final-table.Rmd

• that annotates each Fusion found in a sample with kinase domain retention status and overexpression status of BRAF in the fusion
• filters for the following condition:
  a) kinase domain retention == "Yes"
  b) Fusion_Type=="in-frame" ( so that if kinase domain is retained its coded in the fusion transcript?)
  c) overexpression of BRAF == "Yes"
• and outputfile will have column names compatible with pbta-fusion-putative-oncogenic.tsv to be able to run through 01-make-lgat-final-table.Rmd successfully.

What input data should be used? Which data were used in the version being updated?

• pbta-fusion-putative-oncogenic.tsv from v16
• external data from pfam and biomart will need to be added and modified to get domain level information and coordinates

When do you expect the revised analysis will be completed?

1 week

Who will complete the updated analysis?

@kgaonkar6

[kgaonkar6]

I looked into the zscores using this function which doesn't show consistent overexpression of BRAF across the BRAF fusions.
source(file.path("..","focal-cn-file-preparation","util", "rna-expression-functions.R")) for expression scores for all BRAF fusion in lgat samples:

When using the Gtex zscore from fusion_filtering analysis it shows all overexpression. If it's Gene1B it does look like on average the zscore is higher than if BRAF it's in Gene1A . Using this code it also seems that all fusions with gene1A ==BRAF doesn't retain domain.

[jainpayal022]

Can help address by Friday

[kgaonkar6]

https://www-sciencedirect-com.proxy.library.upenn.edu/science/article/pii/S1535610820301513?via%3Dihub
they've looked into KIAA1549--BRAF fusion with different breakpoints ( exon16 -> exon9) ,(exon15 -> exon10) and (exon15 -> exon9) and briefly discussed that non-KIAA1549 BRAF fusions occurred exclusively in adolescents with no progression events

Because of the large number of tumors, we could sub-stratify the BRAF fusions into subgroups based on their breakpoints (Figure S4A). The most common KIAA1549-BRAF fusion involved exon 16 in KIAA1549 and exon 9 in BRAF (16:09). Like all KIAA1549-BRAF fusions, 16:09 was significantly enriched in pilocytic astrocytoma (n = 73/83, 88%, p < 0.0001, Fisher's exact test) and in cerebellar tumors (n = 60/83, 72%, p < 0.0001, Fisher's exact test) (Figures S4B and S4C). Interestingly, the only KIAA1549-BRAF fusion seen in hemispheric tumors was 15:09 (Figure S4B). 15:09 was also the primary fusion seen in tumors with extensive dissemination (n = 5, 83%, p < 0.0001, Fisher's exact test). 15:09 fusions were associated with a worse PFS (p = 0.0003, log rank test, Figure S4D), with a 5-year PFS of 59% as compared with 77%–100% for other fusion subtypes. BRAF fusions not involving KIAA1549 occurred exclusively in adolescents with no progression events (median follow-up = 3.7 years), while 15:11 was only observed in two infants who rapidly experienced tumor progression and died (Figures S4D and S4E). Identifying the specific fusion breakpoints of KIAA1549-BRAF will be important to properly ascertain their propensity for certain clinical features and impact on outcome.

[jharenza]

@kgaonkar6 let's revisit this with #790, in which I am adding additional BRAF fusions as another subtype. I think the point you raise about the kinase domain retention is a great one, and we should consider this.

[jharenza]

closing this as #790 will cover this