Bernard Kim Profile
Bernard Kim

@Bernard_Y_Kim

892
Followers
524
Following
46
Media
374
Statuses

Postdoc in @PetrovADmitri 's lab @Stanford . Interpreting popgen processes of Drosophila at macroevolutionary time scales. @lohmueller lab alum

San Francisco, CA
Joined September 2012
Don't wanna be here? Send us removal request.
Pinned Tweet
@Bernard_Y_Kim
Bernard Kim
4 months
Very stoked to share that I am joining @Princeton EEB as an Assistant Professor in Jan 2025. If studying evolution using genomes AND popgen data from 100s of species sounds interesting, do reach out! I’ll be recruiting at all levels for work spanning field to wet lab to comp bio.
Tweet media one
Tweet media two
43
54
397
@Bernard_Y_Kim
Bernard Kim
3 years
An bit of an unexpected finding for @PetrovADmitri , @danrdanny and I: with a few minor tweaks, @nanopore is great at assembling genomes from single drosophilids collected from the wild (collections w/or courtesy of @m_karageorgi , @DarrenObbard , Thomas Werner, Don Price, et al).
Tweet media one
5
25
116
@Bernard_Y_Kim
Bernard Kim
2 years
@PetrovADmitri and I have been working with amazing Drosophila collaborators to broadly sequence the diversity of this group. We believe it is now technically feasible to sequence all >4,400 species, and have organized a #Dros23 workshop to engage the community in this effort.
Tweet media one
5
32
113
@Bernard_Y_Kim
Bernard Kim
2 years
So impressed by our first @nanopore PromethION runs (P2 solo) prepped by @kasolari and @_hgellert . Zebra blood gDNA + ligation kit 14 with minimal tweaks. Barcoding workflow needs tweaks for read length but getting tons of data doing 8 Drosophila samples at a time.
Tweet media one
3
9
56
@Bernard_Y_Kim
Bernard Kim
2 years
If anyone's interested in running single-drosophilid assemblies themselves, version 1 is finally up on . Please note we are still working out some kinks and optimizing. Only tested with @nanopore MinION and LSK110.
@Bernard_Y_Kim
Bernard Kim
3 years
An bit of an unexpected finding for @PetrovADmitri , @danrdanny and I: with a few minor tweaks, @nanopore is great at assembling genomes from single drosophilids collected from the wild (collections w/or courtesy of @m_karageorgi , @DarrenObbard , Thomas Werner, Don Price, et al).
Tweet media one
5
25
116
0
27
50
@Bernard_Y_Kim
Bernard Kim
5 years
@ER_Ebel and I have been running @nanopore LSK109 preps but where size selective precipitation with @circulomics SRE kit is used in place of bead cleanups. Protocol needs tweaking to improve yield of ultra-long reads but already giving great results. Longest reads ~630kb.
Tweet media one
4
14
45
@Bernard_Y_Kim
Bernard Kim
3 months
We're incredibly excited for our own interests, but perhaps more importantly hope to build something for the entire scientific community to use. We continue to work on 100s more species - even outside Drosophilidae - and will always release data with zero restrictions.
@PetrovADmitri
Dmitri Petrov
3 months
Thrilled about the possibilities this opens up. We aim to sequence genomes of O(all) Drosophila species with mult. individuals per species to combine pop. genomics with mol. evolution in a single system and to generate high resolution estimates of selection per gene or even codon
7
48
181
0
7
30
@Bernard_Y_Kim
Bernard Kim
4 years
@PetrovADmitri Thanks Dmitri (and everyone). I think it speaks more to the quality of writing advice received from you, @MollySchumer , and labmates. Also very lucky to be working with such phenomenal and inspiring collaborators.
4
1
23
@Bernard_Y_Kim
Bernard Kim
2 years
To learn more, please come to our workshop from 7:45-9:45PM Friday at #Dros23 ! We will have additional broad-scale perspectives on biodiversity, a Drosophilidae Tree of Life @AntonySuvorov , clade-scale TE annotation @GonzalezLab_BCN , and drosophilid microbes @drosobachia
Tweet media one
0
9
15
@Bernard_Y_Kim
Bernard Kim
5 years
Our hybrid @nanopore ligation kit protocol using @circulomics SRE in place of beads is slowly getting better. Average read length is 16.5kb this run, and we have a nice tail of reads >100kb. It would've been better but it appears we have a bad flow cell (although it QC'ed fine).
Tweet media one
Tweet media two
Tweet media three
2
3
15
@Bernard_Y_Kim
Bernard Kim
6 years
@lohmueller @PetrovADmitri I couldn't have done any of it without an advisor and lab that had my back, though. Thanks for everything, Kirk!
1
0
15
@Bernard_Y_Kim
Bernard Kim
5 years
Looks like precipitation preps are the key to big amounts of long reads
@DrT1973
John Tyson
5 years
Pushing for 100Kb+ reads with a modified @nanopore LSK109 ligation prep. Gone bead-free and old school with PEG/NaCl and a shearing tickle. Sure makes things simpler, cheaper and longer 😉. @longreadclub
Tweet media one
13
77
224
0
3
14
@Bernard_Y_Kim
Bernard Kim
4 months
Looking forward to presenting our work at this meeting, and grateful for the opportunity to connect with the Brazilian Drosophila community!
@segedxii
SEGED XII
4 months
Dr. @Bernard_Y_Kim estará na sessão de Evolução. Leia mais sobre os palestrantes no nosso site (link na bio) na aba PALESTRANTES. Dr. @Bernard_Y_Kim will be at the Evolution session. Read more about the speakers on our website (link in bio) in the SPEAKERS tab. #SEGEDXII
Tweet media one
Tweet media two
0
0
5
0
1
12
@Bernard_Y_Kim
Bernard Kim
2 years
We've also performed duplex calling on both 260/400bps datasets to get ~13X depth of duplex/Q30 reads. Duplex and simplex (reads used for duplex removed) calls available via BioProject PRJNA914057. ~20% of data had IDed pairs, so coverage of duplex reads is ~10% of original data.
@nanopore
Oxford Nanopore
2 years
As a first Community collaboration the Open Datasets project now hosts raw sequencing device data (fast5 files) kindly provided by @Bernard_Y_Kim at Stanford University from a Drosophilae Melanogaster strain. Read here: 2/3
Tweet media one
1
0
3
0
2
11
@Bernard_Y_Kim
Bernard Kim
5 years
Multiple (4-5) library loads are prepared with 0.5X-1X volume of ligation kit reagents. Determine vol library to use, mix 1:1 library:SQB, then add flush buffer (must be from EXP-FLP002, unmixed with FLT) to a final volume of 80uL. No LB. E.g.: 10 uL library, 10 uL SQB, 60 uL FB.
1
1
11
@Bernard_Y_Kim
Bernard Kim
2 years
@PetrovADmitri With just @nanopore sequencing, we are capable of generating a draft assembly from one wild-caught fly for $100. Illumina prices have also dropped to <$1 per 1X coverage of a Drosophila genome.
1
2
11
@Bernard_Y_Kim
Bernard Kim
5 years
This is a 70 hour old flow cell after a DNAse flush.
Tweet media one
1
0
10
@Bernard_Y_Kim
Bernard Kim
2 years
@_ellie_cat @_hgellert and I will be working on hundreds of new drosophilid genomes over the next 12 months - many in interesting phylogenetic positions. Please reach out if you are interesting in collaborating!
2
0
10
@Bernard_Y_Kim
Bernard Kim
5 years
@PetrovADmitri @RoshniAPatel oh no, that reminds me of this moment... 🤦
Tweet media one
1
0
10
@Bernard_Y_Kim
Bernard Kim
4 years
@Jazlyn_Mooney This juvenile and sexist behavior is just... wow. I'm appalled that you had to be on the receiving end of it. Glad we know who to avoid in the future though.
0
0
8
@Bernard_Y_Kim
Bernard Kim
4 years
17/17 For now we’d just like to provide these genomes as a resource for the community, and we are looking forward to sharing our future work with you in the same open manner. Send me a DM if you want to stay in touch through our Slack workspace.
3
1
8
@Bernard_Y_Kim
Bernard Kim
4 months
@Jazlyn_Mooney @Princeton i'm really excited to finally join all the @lohmueller alum ( @arundurvasula @christiandhuber @dortega_delv et al.) as a prof!
0
0
7
@Bernard_Y_Kim
Bernard Kim
6 years
@random_hike @jgschraiber @RyanGutenkunst @lohmueller sorry i'm late to this party, but i wrote up a brief "how-to" document for fitting a DFE that tries to be clear about these technical issues.
1
1
7
@Bernard_Y_Kim
Bernard Kim
4 years
4/17 With recent advances in long read (particularly @Nanopore ) sequencing genome assembly has become a lot easier. We want to specifically highlight something from @danrdanny ’s 2018 paper: a remarkable cost benchmark of US $1,000 per genome.
1
2
7
@Bernard_Y_Kim
Bernard Kim
6 years
Congrats Jazlyn!
@Jazlyn_Mooney
Jazlyn Mooney, PhD
6 years
My pre-print on patterns of genetic variation in Latin American Isolates is finally up! We find that the demographic history of population isolates strongly influences geomic patterns of variation, and there is no single signature of a population isolate.
3
21
48
1
0
7
@Bernard_Y_Kim
Bernard Kim
3 years
@pastramimachine Microsoft is crushing it lately, IMO significantly cheaper & more usable for personal scientific computing than Mac or straight Linux. Our community hasn't really picked it up though (yet). WSL, GPU support, Docker Desktop + NVIDIA docker mostly working, now GUI support...
1
0
7
@Bernard_Y_Kim
Bernard Kim
2 years
Even with our currently limited sampling we are seeing some very cool results. Here you see that the same genes experience high rates of adaptive substitution in different species groups!
Tweet media one
2
0
7
@Bernard_Y_Kim
Bernard Kim
4 years
@DavidEnard This is one of the best things about the new SLiM GUI. Students/interns no longer need an overpriced laptop with a horrible keyboard to use this kick ass program. This only took a couple hours of googling to figure out. Credit should go to the SLiM team for making it happen.
0
0
7
@Bernard_Y_Kim
Bernard Kim
2 years
@nanopore @PetrovADmitri @danrdanny @danielrmatute @DarrenObbard @_hgellert @mbeisen The issue of indels in protein-coding exons seems to largely be solved. Here we have excluded genes where we consistently find large (100s of bp) CDS indel differences between the dm6 reference and long-read (ONT and PacBio) iso-1 assemblies.
Tweet media one
1
2
7
@Bernard_Y_Kim
Bernard Kim
3 years
Quite excited for the popgen/phylogenomics/etc. possibilities this opens up & aiming to have finished sequences available on NCBI in a few months, along with open-source protocols.
1
0
6
@Bernard_Y_Kim
Bernard Kim
4 years
16/17 We are always looking to collaborate so please reach out if you’re interested! We’re searching for more genomes to add to the alignment and polymorphism from additional species. We plan to release this as a community resource too.
1
0
6
@Bernard_Y_Kim
Bernard Kim
1 year
@nanopore We’ve done lots of fresh collections of wild flies over the last couple years: across the Hawaiian Islands, the Western US (CA, OR, WA, MT, ID, CO), the Midwest (MI, OH), Europe (UK) and were able to assemble single-fly genomes for every species we collected.
Tweet media one
Tweet media two
Tweet media three
1
1
6
@Bernard_Y_Kim
Bernard Kim
3 years
PS if anyone has similar drosophila samples (wild, etoh, no reference) & interested in collaborating on similar attempts at assembly please let us know!
2
1
6
@Bernard_Y_Kim
Bernard Kim
1 year
@nanopore This effort is being managed as an open science, community resource. Data are on NCBI, containerized+Snakemake pipelines on GitHub, public protocols, and a 298-way Cactus alignment is available for download. Please see the preprint for more details on accessing these resources.
1
1
5
@Bernard_Y_Kim
Bernard Kim
5 years
Extremely grateful to @DrT1973 , @danrdanny , and @circulomics for contributing their assistance moving this forward. They have been incredibly generous with their time.
0
1
6
@Bernard_Y_Kim
Bernard Kim
4 years
@PetrovADmitri @PravruthaRaman @HarmitMalik been out of cell reception but we have nanopore & illumina sequences of biarmipes, subpulchrella, and takahashii and happy to share
1
1
6
@Bernard_Y_Kim
Bernard Kim
5 years
The rough workflow is now: 1) Circulomics SRE 2) Repair and end-prep 3) Circulomics SRE 4) Adapter ligation 5) Circulomics SRE Trying to get better consistency in the preps but will post a detailed protocol once those kinks are ironed out.
0
0
6
@Bernard_Y_Kim
Bernard Kim
5 years
Also @circulomics team has been super helpful for working together to figure all this out.
0
0
6
@Bernard_Y_Kim
Bernard Kim
4 years
13/17 This was a community effort with many groups providing resources and flies. We think the right thing to do, particularly with data generated like this, is to embrace open science principles and release the dataset as a community resource for everyone to dig into.
1
0
5
@Bernard_Y_Kim
Bernard Kim
4 years
@petrelharp FWIW, Microsoft is reportedly bringing native X11 windowing and GPU support to Windows Subsystem for Linux later this year. Then running SLiM GUI across any system should super easy!
1
1
5
@Bernard_Y_Kim
Bernard Kim
5 years
@schwessinger @DrT1973 @nanopore @circulomics @protocolsIO Sure, happy to post a protocol soon. Still trying to get things tweaked to improve yield of ultra longs. The other thing I want is to make sure its robust enough to reproduce results across multiple samples without too much fiddliness. Imagine it would be more useful that way.
1
1
5
@Bernard_Y_Kim
Bernard Kim
4 years
14/17 There’s so much to be done. We ( @AAComeault @VictoriaBelle18 ) are incorporating these and many other genomes into a large progressive Cactus (whole-genome) alignment of ~200 genomes, including different versions/strains for some species. See this guide tree for a preview.
Tweet media one
1
1
5
@Bernard_Y_Kim
Bernard Kim
6 years
@lohmueller More than anything, I'm lucky to have such great mentors, including @random_hike . Hopefully I can continue that tradition when I join @PetrovADmitri 's lab.
0
0
5
@Bernard_Y_Kim
Bernard Kim
4 years
9/17 By using these techniques, we were able to significantly increase the biological, geographic, and phylogenetic diversity of drosophilid genomes available to the scientific community, all in a cost-effective manner.
1
0
4
@Bernard_Y_Kim
Bernard Kim
2 years
Through the invaluable contributions of field collections via Don Price, Patrick O'Grady, Masanori Toda, Thomas Werner, @DarrenObbard , and many other amazing collaborators, we have population genomic datasets across >100 species and counting.
Tweet media one
1
0
4
@Bernard_Y_Kim
Bernard Kim
1 year
@nanopore This approach also worked surprisingly well for ethanol-preserved specimens, opening older ethanol preserved specimens for genomic study, helping us add many esoteric taxa to the tree.
Tweet media one
1
1
4
@Bernard_Y_Kim
Bernard Kim
1 year
@nanopore @ASRampasso @KylaOHearn @erdagostino @_LadyOfTheFlies @afmata_2 @Ldy_Natalia @YazBraimah Also big thanks to Patrick O'Grady, Dan Barbash, @ClarkLabCornell , @mbeisen , @PetrovADmitri , @_hgellert , @TheAyrolesLab , and others for all their help with organizing and otherwise making this happen.
0
0
4
@Bernard_Y_Kim
Bernard Kim
2 years
@TJesse62 @nanopore @PetrovADmitri @danrdanny @danielrmatute @DarrenObbard @_hgellert @mbeisen Nope, I thought that was curious too. We consistently see mean QV drop slightly with Illumina polishing but median QV (genomic windows) increases, greatly in CDS. Our downstream work uses synonymous/nonsynonymous variation (lowest hanging fruit) so optimizing CDS is top priority.
1
0
4
@Bernard_Y_Kim
Bernard Kim
4 years
3/17 Drosophila were among the first groups for which genomes of multiple species (including old and young divergences) were assembled. Early genomic data was limited to a few species/groups, a consequence of how difficult/costly it was to create a reference quality genome.
1
0
4
@Bernard_Y_Kim
Bernard Kim
6 years
Congrats Arun! Looking forward to reading this.
1
0
3