Banca de QUALIFICAÇÃO: KARLA CRISTINA TABOSA MACHADO

Uma banca de QUALIFICAÇÃO de DOUTORADO foi cadastrada pelo programa.
STUDENT : KARLA CRISTINA TABOSA MACHADO
DATE: 31/07/2020
TIME: 09:00
LOCAL: Via Google Meet devido pandemia covid
TITLE:

Metaproteomics of human tissues to identify molecular targets of clinical interest

KEY WORDS:

Proteomic, meta-analyses,  biomarker, antigens, cancer/ testis

PAGES: 59
BIG AREA: Ciências Biológicas
AREA: Bioquímica
SUBÁREA: Química de Macromoléculas
SPECIALTY: Proteínas
SUMMARY:

Next-generation sequencing technology development made possible to obtain a better characterization of the human genome transcriptome. Similar technological breakthroughs also happened for proteomics in the last decade regarding sensitivity and throughput, but not at the same level as genomics. The profile of the whole tumor environment was often based on DNA and RNA analysis. But genomic and transcriptomic studies are not sufficient to elucidate all molecular mechanisms in the cell, since function occurs mostly at the protein level. Moreover, the amount of mRNA is not necessarily proportional to the translated protein level. Advancements in proteomic approaches allowed to use the proteome to explore cancer molecular characterization, as well as to reveal new biomarkers, leading toward personalized medicine. One difficulty is that while transcriptomics studies can be done using hundreds of samples from cells or tissues, proteomics studies work with few samples. To solve this problem, this work suggests the integration of proteomic data from different projects available in public repositories, allowing a more comprehensive view of the samples under analysis. This work aims to perform a computational meta-analysis of proteomic data from human tissues, in order to identify proteins of clinical interest for cancer. Approximately 10 Tb of proteomic data were analyzed, containing more than 500 samples of healthy tissues, tumors collected from patients and immortalized cell lines used as a model in cancer. After integrating the proteomic data, the samples were grouped according to the tissues to which they belonged, for purpose to amplify the sample number by tissue. Such clustering revealed 369 tumoral samples obtained from 8 different types of cancer. The healthy tissue samples were also regarded, totaling 140 samples. Principal Component Analysis (PCA) was used to analyze protein expression in samples. Notably, it was observed that clustering  was not efficient strategy for breast cancer and melanoma, possibly due to the heterogeneity of these tumors. Finally, to perform the identification of proteins as molecular targets of clinical interest, the subsequent analysis focused on at least three aspects: cancer / testis antigens (CTAs) characterization, transcription factors and upstream ORFs. So far, only CTAs have been analyzed. To identify which CTAs were expressed at the protein level, CTAs previously predicted in transcriptomics works were used. As a result, 222 CTAs were identified at the proteomic level, of which 19 were differentially expressed in tumoral samples compared to healthy sample. In conclusion, the computacional meta-analysis performed in this study to show potential to enable future advances in the characterization of tumoral metaproteomic and, consequently, in the identification of new protein biomarkers for the treatment against cancer.

BANKING MEMBERS:
Interno - 1267860 - GUSTAVO ANTONIO DE SOUZA
Interno - 1507794 - RODRIGO JULIANI SIQUEIRA DALMOLIN
Interno - 1939184 - SANDRO JOSE DE SOUZA