Construction of evolutionary trees

TS Tomoko Saito
AN Atsushi Niida
RU Ryutaro Uchi
HH Hidenari Hirata
HK Hisateru Komatsu
SS Shotaro Sakimura
SH Shuto Hayashi
SN Sho Nambara
YK Yosuke Kuroda
SI Shuhei Ito
HE Hidetoshi Eguchi
TM Takaaki Masuda
KS Keishi Sugimachi
TT Taro Tobo
HN Haruto Nishida
TD Tsutomu Daa
KC Kenichi Chiba
YS Yuichi Shiraishi
TY Tetsuichi Yoshizato
MK Masaaki Kodama
TO Tadayoshi Okimoto
KM Kazuhiro Mizukami
RO Ryo Ogawa
KO Kazuhisa Okamoto
MS Mitsutaka Shuto
KF Kensuke Fukuda
YM Yusuke Matsui
TS Teppei Shimamura
TH Takanori Hasegawa
YD Yuichiro Doki
SN Satoshi Nagayama
KY Kazutaka Yamada
MK Mamoru Kato
TS Tatsuhiro Shibata
MM Masaki Mori
HA Hiroyuki Aburatani
KM Kazunari Murakami
YS Yutaka Suzuki
SO Seishi Ogawa
SM Satoru Miyano
KM Koshi Mimori
request Request a Protocol
ask Ask a question
Favorite

From the multiregion sequencing data for each case, an evolutionary tree was constructed using the Treeomics algorithm21 (https://github.com/johannesreiter/treeomics) with default parameters. For every mutation existing in the multiregion mutation profile, the numbers of variant reads, read depth, chromosomal coordinates, gene symbol, and substitution pattern were prepared as input data to Treeomics. Treeomics not only constructs an evolutionarily tree but also corrects potential sequencing artifacts so that all mutations have mutation patterns compatible with the topologies of the evolutionary tree. Based on the parts of the tree that the mutations constituted, we obtained trunk, branch, internal branch, and external branch mutations, which were refined versions of ubiquitous, heterogeneous, shared and private mutations, respectively. To remove potential sequencing artifacts, Treeomics also employs mutation filters, which filtered out 1.3% of our input mutations. Information about the trunk-branch categorization is also provided in Supplementary Data 3. We were unable to apply Treeomics to the ACRC3 data, which contained 21 samples, due to insufficient memory on our computer. To address this problem, we divided the ACRC3 data into two parts which corresponded to two apparent sample clusters in the multiregion profiles. After the divided data were subjected to Treeomics, an evolutionary tree was constructed by merging the results. Except for ACRC2 and ACRC3, the robustness of the evolutionary tree inference was examined on 1000 bootstrapping samples from the input mutations. For ACRC2 and ACRC3, only 50 bootstrapping samples were obtained due to the memory limitation. The inferred evolutionary trees were annotated with the same driver gene list as used for the heat maps of the multiregion mutation profiles (Supplementary Table 1). For detection of subclonal mixing, we reconstructed evolutionary trees with the “-u” option and the obtained information of subclonal mixing was added to the trees constructed without the “-u” option (Supplementary Figs. 6 and 7).

Do you have any questions about this protocol?

Post your question to gather feedback from the community. We will also invite the authors of this article to respond.

post Post a Question
0 Q&A