Latest News
DeMixSC paper is published at Genome Research!
DeMixSC, an innovative deconvolution approach that overcomes the technological discrepancy between bulk and sc/snRNA-seq data using an improved wNNLS framework. It achieves high accuracy and generalizability, requiring only a small tissue-matched benchmark dataset for the targeted large bulk cohorts.
Free access at https://genome.cshlp.org/content/35/1/147
Exciting Updates from Stats Up AI
We’re thrilled to share two major milestones for Stats Up AI:
1.ASA Approval:
The Stats Up AI interest group has been officially approved by the
American Statistical Association, recognizing its mission to foster
collaboration and innovation at the intersection of statistics and AI.
2.Harvard Data Science Review Publication:
Our recent event, Stats and AI- A Fireside Conversation, was a great
success. A summary of the discussion has been accepted for publication
in the Harvard Data Science Review, further amplifying its impact.
These achievements highlight the growing influence of Stats Up AI in
advancing
the integration of statistics and AI. Stay tuned for future updates as we
continue to engage with this inspiring community!
Congratulations to Carissa for winning the ABRCMS Presentation Award!
Congratulations to our summer intern, Carissa Fong, for winning the presentation award at ABRCMS (Annual Biomedical Research Conference for Minoritized Scientists)! Her award-winning presentation showcased machine learning approaches to effectively predict tumor-specific mRNA expression (TmS). We are so proud of her achievement!
Wang Lab Postdoc Ankita Paul Awarded MD Anderson IDSO Fellowship
Congratulations to our postdoc Ankita, who has been awarded the MD Anderson Institute for Data Science in Oncology (IDSO) Fellowship! This fellowship is a great opportunity that provides junior researchers with advanced training in applying data science to oncology. We are excited to see the impactful contributions she will make through this program!
Wenyi Hooded PhD Graduates Nam and Yunjie(Jeffery) at Rice Univerisity Commencement
Congratuations again to Dr. Nguyen and Dr. Jiang. Wish you all the best and may your future endeavors be filled with success and fulfillment!
MuSE2.0 paper is online at Genome Research online!
We are exctied to officially introduce MuSE2.0, which reduces computing time by up to 50x compared to MuSE 1.0 and 8-80x compared to other popular callers. Our benchmark study suggests combining MuSE2.0 and the recently accelerated Strelka2 achieves high efficiency and accuracy in analyzing large cancer genomic datasets.
Free access here
Journal of Clinical Oncology paper is online!
We conducted a validation study of our LFSPRO software suite, which was developed for risk predictions in families with Li-Fraumeni syndrome, on a clinical patient cohort collected as part of the Clinical Cancer Genetics program at MD Anderson Cancer Center. Unlike research datasets that are meticulously collected over decades for research purposes, our unique dataset closely resembles what genetic counselors observe in real counseling sessions. The validation results indicate that our risk prediction models have the potential to assist decision making in clinical settings, and further highlight the importance of such validation in bridging the gap between methodology research labs and clinics.
Free access at https://ascopubs.org/doi/10.1200/JCO.23.01926
Congratulations to Dr. Nguyen and Dr. Jiang!
Congratulations to Nam H Nguyen and Yujie Jiang for successfully
defending
their PhD theses and earning their doctorates!
Don't miss out! Join the Upcoming Webinar: "Empowering Statistics in the Era of AI - A Fireside Conversation"
STATS UP AI - A community for Statistics and Biostatistics
Professors from multiple universities initiate StatsUpAI, aiming
to
elevate
the role of statisticians in AI research.
This movement emphasizes empowering statisticians to lead and
innovate in
addressing real-world challenges through AI research.
As one of the founding members, Wenyi's involvement highlights
her
commitment to advancing statistical methodologies in the era of
artificial
intelligence.
To learn more about StatsUpAI, visit their webpage at
https://statsupai.org
LFSPROShiny paper is published online!
LFSPROShiny is an interactive R/Shiny application designed to
perform risk
prediction and visualization for Li-Fraumeni syndrome (LFS),
a genetic disorder associated with TP53 mutations, enabling
genetic
counselors to assess patient risk profiles and support informed
decision-making without the need for
programming expertise.
Free access at
https://ascopubs.org/doi/10.1200/CCI.23.00167.
Join us on the 2023 Leading Edge of Cancer Research Symposium hosted by MD Anderson on Nov 16-17!
This in-person event provides an incredible no-cost opportunity
to
engage
with and
learn from national and international leaders in cancer
research,
including
an
opportunity to present new research at our poster session.
Ten of the top posters will be chosen for presentations as part
of
the
symposium
agenda as well as monetary awards. Deadline to submit an
abstract is
Oct
20. Please take your lab and program trainees
to
join us!
Click for more details.
Congratulations to the launch of Institute for Data Science in Oncology(IDSO) at MD Anderson!
IDSO integrates the most advanced computational and data science
approaches
with the
institution’s extensive scientific and clinical expertise,
aiming to
profoundly
enhance patient lives by revolutionizing oncological care and
research.
Dr. Wang's lab is proudly affiliated with this pioneering
initiative,
dedicated
to advancements in cancer care and research through innovative
methodologies.
DeMixSC paper is on bioRxiv!
The difference in technology between bulk and sc/snRNA-seq data
significantly
diminishes the accuracy of current deconvolution methods. To
address
this
issue,
we've introduced DeMixSC,
an innovative deconvolution approach that overcomes this
challenge
using an
improved
wNNLS framework. DeMixSC is distinguished by its accuracy in
deconvolution
and its
generalizability to be applied to any large bulk cohorts.
All it requires is a small set of benchmark dataset that match
the
tissue-type of
the targeted large bulk cohorts.
Click for software tool and paper preprint.
