bioinformatics for multiomeโspatialโsingle-cell; developmentโdisease; Assistant professor at UCSF; Tweets๐own; BlueSky: PengHe.bsky.social Mastodon:penghe
San Francisco, CA
- Tweets 1,066
- Following 1,709
- Followers 1,306
- Likes 1,020
Pinned Tweet
Our Perspective paper "Toward informed batch correction for single-cell transcriptome integration" is out in @NatComputSci ๐rdcu.be/e4cSp
We review a decade of batch-correction methods and propose to move from "blind" integration to "informed" modeling.๐งต๐
1
4
12
1,488
๐จ๐๐ ๐๐๐ job opportunity @UCSF ๐จ:
I'm hiring TWO postdocs - an experimentalist and a computational biologist. Join us to use cutting-edge single-cell and spatial technologies to unravel the mysteries of human tissues.
opportunities.ucsf.edu/conteโฆ
1
91
193
32,375
๐My first postdoc project at @teichlab is now published online! Huge thanks to @ZHANGLAB, Bao Zhang, John Lawrence, @Shuaiyu_Wang, and our collaborators for making this exciting research possible. nature.com/articles/s41586-0โฆ Here are 8 of my favorite highlights from the paper:
9
27
154
23,609
Thanks to my PhD supervisor Barbara Wold at @Caltech and my wet lab collaborator Brian Williams, as well as all my co-authors @AxelVisel @LenPennacchio @rosshardison Renlab et al. and friends who supported this piece of work. nature.com/articles/s41586-0โฆ
9
22
126
Thrilled to share our human fetal lung multi-omic atlas to you ๐บ๏ธ๐ผ๐ซ with my comrades @KyungtaeL & @sundakao, a collaboration among 3โฃ๏ธlabs (@LabRawlins @MarioniLab @teichlab) supervised by Emma Rawlins and Kerstin Meyer #HumanFetalLung2022 biorxiv.org/content/10.1101/โฆ
7
22
114
Our preprint is out!
The changing mouse embryo transcriptome at whole tissue and single-cell resolution biorxiv.org/content/10.1101/โฆ
10
26
84
Happy to share our preprint ๐ single-cell + spatial atlas of the developing human limb ๐ฆต๐คณ produced by @teichlab .Thanks to all of our co-authors and friends! biorxiv.org/content/10.1101/โฆ
7
26
85
๐My third project during postdoc (#LungImmuneAtlas), a collaboration with brilliant scientists at @drmarkonikolic's lab (especially @DrJoBarnes & @masahiroy15) and others in Cambridge and London was published in @SciImmunology last Friday!๐งตโฌ๏ธThread here sanger.ac.uk/news_item/immunโฆ
4
10
63
12,184
Iโve opened two more positions in my lab for both junior and senior RAsโno PhD required. ๐ Spread the word!
๐ฌ Dry Lab: aprecruit.ucsf.edu/JPF05334
๐ง Wet Lab: aprecruit.ucsf.edu/JPF05332
Thanks for sharing!
3
33
64
12,161
Happy to share our new preprint about how immune cells can modulate lung epithelium development, produced by @drmarkonikolic's lab and @teichlab biorxiv.org/content/10.1101/โฆ
6
8
43
6,512
I'm maintaining an active list of biology-related job openings at various levels (PIs, staff scientists, postdocs, PhDs, etc.). Please feel free to subscribe or contribute! github.com/brianpenghe/Bio-Jโฆ
1
10
39
4,444
Thank you all for coming to my talk and participating in the inspiring discussions!
You wonโt want to miss Dr Peng He (Wellcome Sanger Institute, Cambridge) presenting โSpatial and multi-omic studies of human organogenesisโ as part of our external seminar series today! @PengHeCam
25th Oct
12-1pm
Open to all Oxford University members.
talks.ox.ac.uk/talks/id/c0f5โฆ
37
The developing human lung has been mapped in unprecedented detail, identifying 144 cell states in the early stages of life, and uncovering new links between developmental cells and lung cancer.
ebi.ac.uk/about/news/researcโฆ
5
5
37
Master #CellxGene in just 13 minutes! Check out this tutorial I created with TikTok and ChatGPT, using our #HumanFetalLung2022 data portal. ๐งฌ
piped.video/3BZdofyr6us
2
6
31
3,338
The Teichmann lab at @sangerinstitute is recruiting Principal/Senior bioinformaticians. As a Programme, we are committed to personal and professional development and would work with you to support your career aspirations. jobs.sanger.ac.uk/intranet/wโฆ
5
23
Happy to see the final version online! Great visualization tool developed by @bayraktar_lab and led by @BioinfoTongLI #team #Science
2
5
28
4,261
Itโs the 13th day after I acquired COVID virus (probably omicron) at my friendโs birthday party. Iโm negative now and eager to share some personal experiences.
1) Lateral flow tests are useful. See my Day 6, Day 11 and Day 12
6
24
Delving into the RNA world of "epigenetics": Excited to share our new paper on how maternal siRNAs jumpstart the piRNA biogenesis feedback loop - a fun collaboration led by @Yicheng_Luo_YL from @AlexeiAravin's lab. #siRNApiRNA #transposon sciencedirect.com/science/arโฆ
2
19
1,683
Very happy to have contributed to showcasing the ๐ features of this incredible platform ๐, dedicated to making basic single-cell/nucleus and๐บ๏ธspatial transcriptomics analysis accessible to everyone, regardless of programming skills ๐ป!!
We were able to quickly reproduce key findings from @PengHeAtlas of the @teichlabโs new work on human limb development using CZ #CellxGeneโs new spatial features.
Give it a try and, as always, please let us know what you think bit.ly/3WUpgvn
1
19
2,061
had a great time at #ISSCR2022 .Thanks for the support from @ISSCR for my travel and talk here! really learned a lot and connected/reconnected to so many friends ๐ค
1
15
Everyone welcome ๐ Public seminar next week UK time Wed 2 March at 5-6pm by @RonZhu2015 : "Multicellular circuit design in mammalian cells".
News ๐ caltech.edu/about/news/modelโฆ
Message me for Zoom link if interested
Abstract see thread below:
1
2
14
Looking forward to this new adventure of our lab!
We are so excited to say that it's been announced that Dr Sarah Teichmann @teichlab will be joining us as Professor in Stem Cell Medicine! Sarah does amazing work, and we're thrilled to be part of it soon.
More here: stemcells.cam.ac.uk/news/sarโฆ
12
1,788
I should've read this earlier, and so should whoever thinks they are studying GRNs. It's a very beautiful theory proposed more than half a century ago which was mostly correct (except that it is mainly the TFs but not lncRNAs that are the "activators") science.sciencemag.org/conteโฆ
1
14
Great opportunity to share the stories and struggles of my earlier career. Science isnt just about linear glory ! And amazing and wholesome stories by Kerstin! Looking forward to the next session!
CamLS hosted our first career and networking event last Friday with with a focus on academia. We were delighted to have Dr Kerstin Meyer and Dr @PengHeCam who shared captivating stories about their career journeys! Huge thanks our sponsor @TheBgiGroup and everyone who joined us!
14
1,312
Fun dinners with Development Biology professors @MekaylaStorer @LabRawlins and future Immunology professor @AmandaOPhD . English winter can be quite enjoyable! โ๏ธ๐ฝ๐ท
1
15
1,214
Final day of the #EMBOLabLead course choreographed by @LeadSculpt with @EMadissoon @EdoardoDImprima and many more @embl ers!
1
1
10
Our adult story of lungcellatlas.org/ led by @EMadissoon and
@AmandaOPhD is also online now! Yet our journey hasn't stopped as more cell types/states (such as angiogenic tip cells) and their regulatory principles are to be described in follow-up papers!
Whatโs more festive than Jingle Bells? The Single Cells in our Spatial Multiomic Lung Cell Atlas finally out in @NatureGenet today bit.ly/3YyhTZc! Huge thank you the whole team @EMadissoon @teichlab @bayraktar_lab @StatGenomics @ParsySaeb K. Meyer M. Clatworthy ๐๐๐ซ
11
1,321
Happy to see our work led by @t_koromila of the Stathopoulos lab officially online now! Drosophila embryoโs spatial-temporal gene regulation is way more interesting than I would have imagined! doi.org/10.7554/eLife.59610 @eLife
3
13
CZI data portal and cellxgene are really helpful in sharing processed data with collaborators ๐
Delve into the data from 71k+ cells in this multiomic atlas of human lung development via #CZCellxGene by @PengHeCam & colleagues in @Teichlab, @LabRawlins, @MarioniLab + @PDandHDLab as published in @CellCellPress bit.ly/3pgMuhl
1
12
1,741
When we type a password, we should not be shown a ********* but should be given this โถโทโธโนโบโปโผโฝโพ so that we know how many digits we've typed when trying to backspace - just a suggestion
9
Just did an analysis of my past working hours โฐ.
This year's goal: have more proper weekends and holidays ๐๏ธ
2
9
Nature has just released a summary of data resources organized by @humancellatlas that can help us understand COVID-19 !! nature.com/articles/s41591-0โฆ
3
5
Very talented and diligent pair of atlasers! Well deserved!
Very pleased and proud to see our work out in @Nature nature.com/articles/s41586-0โฆ
Also great to be at #HCA2023GM Toronto with @kazukane as it goes live! @humancellatlas
The preprint tweetorial lays out the key findings in detail, but here are the headlines...
1/n
8
1,169
Cool! congratulations @hao_wu_7
scRNA-seq offers a snapshot of transcriptomes but not RNA dynamics. scNT-seq from the Wu lab @hao_wu_7 integrates metabolic labeling and droplet sequencing to perform parallel analysis of newly-transcribed and pre-existing mRNAs. nature.com/articles/s41592-0โฆ
1
2
8
just started a resource list of open positions for biologists on Github. Please feel free to share and contribute ๐ค github.com/brianpenghe/Bio-Jโฆ
1
1
8
991
Copy-pasting a list of genes from Excel into python codes? This website saves you a lot of time by adding ' and commas automatically. seabreezecomputers.com/excelโฆ
2
7
We need a website like this to turn marker genes to cell types!!! #celltypist flagid.org/
1
8
Incredibly useful Bootcamp with energetic new friends! Thanks for organizing and I'm looking forward to the coming days!
6
t-SNE looks so beautiful after seeing too many UMAPs
Our paper describing gene regulation in the fly brain is out now in @Nature! So proud of this one and of the amazing team work with @SaraAibar @ibrahimihsan and @joynismail in the @steinaerts lab! Check it out here go.nature.com/3ETX2EF with below our main findings:
8
1โฃHaving accumulated lots of CD45+ cells from developing human lungs in our freezer, we thawed them for BCR, ฮฑฮฒTCR, ฮณฮดTCR, transcriptome, and surface proteome analysis (CITE-seq) at the single-cell level. science.org/doi/10.1126/sciiโฆ
1
3
7
468
Don't be afraid of failure. The way you fail in science is by not being ambitious enough - Jason Chin #EMBLCareersWebinar
1
1
6
I just realized that my "DeepTree" algorithm does something similar to AutoGeneS - select feature genes among highly variable genes.
On the computational side, I developed a simple "DeepTree" algorithm (tutorial here github.com/brianpenghe/pythoโฆ) to select deeply clustered genes from highly variable gene list for single-cell analysis. This can leave behind a lot of batch-effect genes or sporadic noises. (18)
1
7
Interesting discussion thread!
In a new paper with @JanLause & @CellTypist we argue that the best approach for normalization of UMI counts is *analytic Pearson residuals*, using NB model with an offset term for seq depth. + We analyze related 2019 papers by @satijalab and @rafalab. /1
biorxiv.org/content/10.1101/โฆ
5
Congratulations to a Wold Lab alumnus!
Just in! New medicine laureate Ardem Patapoutian and his son Luca, watching the #NobelPrize press conference shortly after finding out the happy news.
Stay tuned for our interview with Patapoutian coming up soon!
Photographer: Nancy Hong
1
7
It'd be interesting to compare with 1. graph-based cell-cycle gene selection method in @teichlab 's thymus paper (science.sciencemag.org/conteโฆ) and 2. hierarchical-clustering based gene selection (nitter.app/PengHeCambridge/statusโฆ) in my #ENCODE3MouseRNA paper. All based on correlation.
Excited to share #DUBStepR, our correlation-based feature selection method for clustering scRNA-seq data! , Link to the preprint: biorxiv.org/content/10.1101/โฆ.
Huge thank you to everyone in @shyam_lab for guiding me through this journey! As always, #tweetorial below :)
6
A fun project about a mysterious mutant and the โepigeneticโ secrets it unlocked. Take a read everyone!
Our new preprint describes how two non-coding RNA pathways cooperate to recognize and repress selfish genetic elements.
Trans-generational cytoplasmic inheritance plays an essential role in this process.
Years of hard work, led by @luo_yicheng.
๐งต /1
biorxiv.org/cgi/content/shorโฆ
1
6
3๏ธโฃWe also figured out a way to computationally stitch Visium data from multiple pseudo-overlapping sections to profile a large piece of tissue (using adjacent slices). Our codes can be found here: github.com/Teichlab/limbcellโฆ
ALT Stitching 3 "pseudo-overlapping" adjacent slices
1
3
5
1,719
Had a great time with @bayraktar_lab
Great talks and discussion at quantitative methods in gene regulation 5, @IOP hq London. #qmgr #biologicalphysics #polnet
1
5
Can't emphasize enough that this is a great platform. Really appreciate the effort of everyone (CZI Cellxgene team and @martinprete for developing the features, as well as my student Weimin Lin and HCA Lattice for formatting the data)
Using the ๐ spatial transcriptomics feature in CZ #CellxGene, we reproduced the key finding of a study from @PengHeAtlas in the Teichmann lab showing transcriptionally + spatially distinct mesenchyme populations in the autopod.
Explore the dataset โก๏ธ bit.ly/4cCLxSW
2
5
819
Finally, all codes are availalbe on our Github repository github.com/Teichlab/lung-immโฆ. Dive into the details and join the scientific discourse! #LungImmuneAtlas
1
6
564
Check out this temporal gene-expression study by @vallens . Nice idea and professional setup : ) piped.video/watch?v=m2nPt1Yhโฆ
5
๐ซ๐บ๏ธ๐
Our spatial lung and airway atlas with multiple novel cell types and an IgA plasma cell survival niche is out in BioRxiv! Excellent multi-lab team effort @Teichlab @bayraktar_lab, Menna Clatworthy, @StatGenomics, @ParsySaeb, Kerstin Meyer.
biorxiv.org/content/10.1101/โฆ
6
Published datasets can still trigger a lot of new hypotheses and discoveries! #Beproactive #datamining #TeamScienceAtItsBest #HumanCellAtlas
New work from the Lung HCA Biological Network: Key nasal cells identified as likely entry points for the new coronavirus. arxiv.org/abs/2003.06122 @humancellatlas #COVID19 #covidใผ19uk
1
5
Earlier this year we had identified a mysterious Foxp1+ perichondrial population in the mouse forelimb nitter.app/PengHeCambridge/statusโฆ Now I'm glad to see it again in human, with more discussions!
Dissecting human skeletal stem cell ontogeny by single-cell transcriptomic and functional analyses biorxiv.org/cgi/content/shorโฆ #bioRxiv
2
6
To stop procrastinating I'll initiate a thread named Lessons I learned from the #ENCODE3MouseRNA project here (0)
1
5
Nice summary!
The third phase of the ENCODE project presents an online registry of over 1,200,000 human and mouse candidate functional elements that regulate genes, and the findings are presented in a collection of papers. See the collection here. go.nature.com/2DlgiRe
6
It's happening! The most popular zoom thesis talk I've seen.
I will be defending my thesis tomorrow, Friday May 27, at 10am PT/1pm ET. All are welcome to join.
Zoom: caltech.zoom.us/j/8708291759โฆ
5
Congrats! ่ๅงๅจๆญฆ!
Congratulations to Sarah Teichmann, who has been elected a Fellow of @royalsociety ๐
Awarded for her contributions to computational biology and #genomics, including her role in founding @humancellatlas
#RSFellows @teichlab
2
5
CTCF definitely cares about orientation. cell.com/fulltext/S0092-8674โฆ
1
5
Real-time cases of confirmed #coronarvirus infection in China can be visualized using the codes developed by @lispczz . I modified their heatmap color scheme (my fork: github.com/brianpenghe/pneumโฆ) to suit my eyes better. Hope the spreads stop soon.
2
2
4
taking a break at Dev 2022 by watching the behaviors of a crayfish outside the seminar room #CoBHumanDev
4
Our institutes need such a place for exhausted scientists / students to sleep. (photo taken 10 years ago at the library of University of Hong Kong)
4
Very interactive and well planned meeting! Highly recommend it for trainees!
A big thanks to #CoBHumanDev organisers, @briscoejames, @priscaliberali, @morris_lab and @xielablife, as well as all the speakers and delegates for making โFrom Stem Cells to Human Developmentโ such a wonderful meeting. We hope to see you all back at Wotton House in 2024!
ALT Delegates from #CoBHumanDev at Wotton House
4
Congratulations @SuYapeng !
Thrilled to see our paper on multiomics of COVID out @CellCellPress This is a great team effort of the worldโs best immunologists &systems biologists @HutchGreenberg @raphg @kickassscience1 @ISBLeeHood @ISBNathanPrice @isbsci @Swedish @Stanford @fredhutch @UCSF @seattlechildren
1
5
I've restarted my Zoom-seminar Google calendar which includes seminar time and Zoom links. If anyone wants to get access to my collection of Genomics seminars please let me know your calendar email address. ๐
5
"genomic features are not i.i.d". Can't agree more.
3. Genomic features are not i.i.d. The genome is SUPER structured. You wouldn't put a promoter in the training set and its gene in the test set, would you? Likewise, the presence of a CTCF motif will tell you about cohesin binding, as well as what ISN'T binding nearby.
4
946
Zic protein family has always been a wonderland for treasure hunt. Opa is one of the most interesting ones, which initiates chromatin landscape changes as a pioneer factor. Nice work @t_koromila and everyone!
Opa (odd-paired) is a zinc-finger transcription factor that controls transition of genes from pair-rule to segmental patterns along the AP axis in D. mel. @t_koromila et al. show it's a late timing factor regulating expression along both the DV & AP axes. biorxiv.org/content/10.1101/โฆ
1
5
Congratulations to @shazanfar @Tim_Lohoff ๐๐๐
So much to read | Integration of spatial and single-cell transcriptomic data elucidates mouse organogenesis | Nature Biotechnology nature.com/articles/s41587-0โฆ
2
5
Indeed! In our #ENCODE3MouseRNA paper I used log(0.1 + x) for bulk RNA-seq analysis and explained to the reviewers that the ratios between 'pseudocount' and 'normalization constant' (1 million in TPM; 10000 in Seurat default for UMI) should be carefully tuned for better results!
5
I was looking for the correct way to extract forward reads (-F 16 is wrong) from bam file, and got an answer by myself half a year ago biostars.org/p/14378/#441049 ๐
4
New year question: 10mM dNTP means how much dATP/dCTP/dGTP/dTTP each in the mix? (just trying to see if I'm the only one who made this serious mistake in grad school)
47%
2.5mM each nucleotide
53%
10mM each nucleotide
95 votes โข Final results
1
5
2,429
1โฃ4โฃ๏ธ
We also learned a couple of lessons. One is that novel clusters may be artifacts, such as contaminations from other organs mapped to other atlases (@EMadissoon 's Esophagus, @cartalop @Mona_Litvinuk โค๏ธ, @roserventotormo placenta and @Muzz_Haniffa 's liver)
1
5
Congratulations!
Delighted to share our study on human skin by โฆ@gary_d_reynoldsโฉ, PeterVegh, JamesFletcher, BethPoyner โฆ@HaniffaLabโฉ. Amazing #teamscience with โฆ@FionaWattMRCโฉ โฆ@teichlabโฉ โฆfor โฆ@humancellatlasโฉ โฆ@wellcometrustโฉ ๐๐พ๐๐พ science.sciencemag.org/conteโฆ
1
4
๐๐๐
I've posted the notes/slides for my computational biology class at github.com/pachterlab/Bi-BE-โฆ
Topics were chosen based on appearing in >=3 bio areas, although for focus examples are all drawn from #scRNAseq.
Homeworks include both theory and exploration of data (via @GoogleColab).
3
1,075
๐Moreover, IL-1ฮฒ-treated lung organoids show a significant increase in ionocytes while IL-13 enhances tuft cells, for reasons we ran out of time to find. Explore our data portal here: fetal-lung-immune.cellgeni.sโฆ
1
4
462
Attending Californian seminars at midnight. Still excited by the work by @KellyStevensLab and @VallierLab
4
1โฃ6โฃ๏ธ
Batch correction may hide important confounded features. I did a โsoft correctionโ by merging highly correlated and highly variable genes per batch. All the codes can be found here. github.com/brianpenghe/pythoโฆ
1
4
Challenge accepted @teichlab and completed with proper social distancing, despite of distraction by the yellow flowers. Fight on Addenbrookes and NHS! Nominating my dearest bros @tomsgoms @laidlawsean to finish a 5K or/and a 5ยฃ donation
1
4
And my exciting June started with an energetic virtual meeting. Thanks to @YuanhuaHuang @XiChenUoM @joshuawkho @linzx06 for organizing this and inviting me to talk about my work. (end of this ๐งต)
So much enjoyed the lively discussion on science in this workshop (locked in HK since moving here). Many thanks to all 42 wonderful speakers & great support from our students and admin colleagues. Such a great experience to organise it with @joshuawkho, @XiChenUoM & @linzx06.
4
