►
From YouTube: DevoWorm (2023, Meeting #18): Segment Cell Tracking, Community Period, Prime Editing and Cell Cycle
Description
GSoC community period: DevoLearn/DW Media/DW Publications. Updates on implementing the Segment Anything Model on the Cell Tracking Challenge dataset. Prime Editing and genetic engineering with precision, the use of biological modules in guiding developmental processes. Cell cycles and transcriptional filtering. Attendees: Sushmanth Reddy Mereddy, Morgan Hough, Jesse Parent, Jyothi Swaroop Reddy, Susan Crawford-Young, Bradly Alicea, and Richard Gordon.
A
B
A
B
B
A
C
A
A
A
Actually
it's
an
organization
on
GitHub,
and
this
is
where
we
do
a
lot
of
the
stuff
with
Devo
learn
and
Devo
graph.
So
we
have
the
Divo,
learn,
repository
and
Evo
graph
repositories
pinned,
and
then
we
also
have
the
models
like
the
nucleus
segmenter
and
some
of
these
others
that
I've
been
worked
on
over
the
application
period.
The
nucleus
segmenter
the
membrane
segment
here
in
the
Woody
age,
population
models
and
all
those
are
sort
of
augmenting
Vivo
learn,
but
will
also
be
useful
for
Devo
graph.
A
We
also
have
some
other
repositories
in
here
that
are
maybe
of
interest
to
people,
so
we
we
have
actually
the
devolar
and
web
is
a
public
archive.
So
we
have.
This
is
where
we
we
had
things
deployed
on
a
web
app,
so
we're
no
longer
doing
that,
but
that
might
be
useful.
Historically,
we
have
some
other
things
here
as
well.
We
have
this
data
science,
demos,
repo,
which
people
in
the
past
have
used
to
publish
their
own
stuff
that
they
worked
on.
A
So
we've
had
people
work
on
Gans
models,
generative
models,
we've
had
people
work
on
Networks
and
other
types
of
centroid
extraction
and
other
types
of
models,
and
then
we
have
these
tutorials
on
data
science
type
stuff.
So
people
doing
things
in
Python,
and
this
is
a
surety
who
pushed
this.
So
this
is
something
I
think
I
remember
this.
This
was
the
neural
style
transfer
thing
that
she
showed
in
one
meeting.
So
we
have
this
in
data
science,
demos.
A
B
C
A
People
have
done
a
lot
of
like
collecting
data
on
individual
cells.
It's
very
hard
to
do
like
it's
not
like
the
type
of
models
we
have
where
we
know
every
cell
and
we
have
a
database
of
every
cell.
We
have
single
cell
data,
but
we
don't
have
I
mean
it's
just
kind
of
like
the
molecular
data
or
I
think
maybe
there's
some
proteomic
data.
C
A
A
C
A
A
A
And
then
first
we
have
the
Devo
worm
repository,
which
has
a
lot
of
the
data
sets
and
a
lot
of
other
types
of
things
like
our
webs,
our
GitHub,
I
o
website
and
other
things.
So
that's
DeVore,
github.com
divawarm
and
that's
a
different
place,
but
this
has
a
lot
of
the
stuff
for
Diva
learns.
So
this
is
probably
the
place
if
you're
interested
in
participating
in
the
community
to
go.
First,
then
I
wanted
to
go
over
last
time.
A
A
A
We
have
media
and
public
lectures,
which
is
actually
interesting,
so
I've
put
together
a
list
of
the
medium
public
lectures
that
we've
done
on
YouTube
and
mostly
on
YouTube
I
guess.
But
these
are,
these
will
go
if
you
go
through
these,
they
should
kind
of
show
the
history
of
the
project
and
some
of
the
different
talks
that
have
been
given
some
of
the
different
projects
that
we
we've
done
and
so
forth.
A
So
the
latest
one
is
something
a
digital
diatoms,
which
is
something
I
prepared
for
a
general
audience
to
kind
of
talk
about
the
work
we've
been
doing
with
diatoms,
with
image
processing
and
with
other
things,
so
I
have
I'm
drawing
from
some
of
the
Publications
we've
put
out,
so
that
should
be,
if
you're
interested
in
what
we're
doing
with
diatoms.
That
should
be
enough
information
to
kind
of
get.
You
started
in
that
direction.
A
B
A
You
know
they're
they're,
not
really
timely,
but
they're
they're
capsules
of
what
we've
done
in
like
certain
years
so
like
in
2022,
this
kind
of
goes
over
what
we
did
that
year.
Other
like
conference
talks
posters.
This
is
a
poster
of
the
mathematics
of
Diva
worms,
so
if
you're
interested
in
the
mathematics
that
we
use
for
divoarm,
this
has
a
number
of
things.
A
So
we
have
things
all
sorts
of
different
types
of
mathematical
Concepts
that
we
deal
with
in
the
project
in
the
group
and
they
kind
of
they're
broken
down
by
category
and
they
kind
of
explain
why.
Why
that's
important?
So
that's
kind
of
an
interesting
approach
to
thinking
about
you
know
like
if
you're
coming
into
the
group.
What
kinds
of
things
are
you
tackling
what
kinds
of
methods
are
you
using
so
forth?
So
that's
that's
a
nice
thing
to
look
at.
A
Let's
see,
yeah,
we
just
have
more
conference
talks,
some
things
on
open
source.
This
is
a
little
older
talk
on
Devo
learn,
so
that's
like
from
21,
so
it
just
kind
of
goes
over
how
it
started
see
yeah.
Then
you
know,
if
you're
interested
in
making
a
tutorial
for
anything
involved
with
the
group
and
what
we
do.
You
know
you're
free
to
do
that
and
usual
Singh.
Who
is
a
gsoc
scholar
in
2020?
A
Did
this
tutorial
for
Diva
learn,
so
he
went
over
how
to
run
different
things
in
the
diva
learn
package
so
that
that
kind
of
serves
as
a
tutorial
but
really
any
kind
of
tutorial.
Even
if
it's
on
something
that's
sort
of
in
a
wide
topic
of
the
group,
so
that
that's
something
that
people
can
do
and
then
we
can
put
that
video.
A
A
So
it's
something
that's
worthwhile.
It
has
a
worthwhile
component
to
it
yeah.
So
we
have
different
talks
here
we
had
we've
had
guest
lectures
too,
and,
and
one
we
had
was
from
George
michaelovsky,
who
did
a
talk
on
structuredness
in
his
idea
of
structural
complexity.
So
he's
actually
done
a
talk
here.
It's
on
our
YouTube
channel.
This
was
back
in
2020.
A
We
also
had
see
we've
had
gsoc
students
and
other
people
give
talks,
and
then
we
we've
also
had
other
people
like
Steve
McGrew,
who's
now
deceased,
who
gave
a
talk
on
on
a
microscopy
method
that
he
worked
on.
So
this
is
a
lot
of
good
stuff
on
this.
In
this
area
we
also
have
worked
on
other
types
of
models
like
cellular
automata
bottles,
and
this
is
a
tutorial
on
something
called
morphozoic.
A
The
morphozoic
is
kind
of
a
forgotten
tool.
It
was
something
that
Tom
portage's
developed.
How
imported
Jesus
is
a
senior
contributor
at
openworm
and
he's
done
a
lot
of
stuff
he's
been
working
with
artificial
intelligence,
since
the
80s
I
believe
so
he's
worked
on
a
lot
of
things.
He's
nowadays
he's
working
on
different
types
of
like
artificial
life.
A
Simulations
and
he's
got
some
really
interesting
ideas.
Well,
anyways.
He
has
this
package
called
morphozoic
and
you
can
actually
you
know
there.
We
have
a
Fork
to
the
repository
on
our
Devo
evil,
worm,
GitHub
account,
and
so
you
go
to
the
diva
worm,
get
out
repository
and
you
can
Fork
it
yourself
or
work
with
it
there,
and
so
he
gave
a
tutorial
in
2016..
Last
week,
I
talked
about
the
open
house
materials
where
we
had.
A
You
know
all
the
projects
from
Diva
or
from
open
worm
in
a
in
a
repository.
One
of
those
was
this
morphozoic
tutorial.
So
in
this
tutorial,
I
I
talked
to
Tom
I
interviewed
him
and
we
walked
through
a
lot
of
the
features
of
morphozoic.
So
he
walks
through
a
tutorial.
He
gives
a
demonstration
and
what's
interesting
about
morphozoic.
Is
that
it's
it
basically
is
able
to
recognize
patterns,
so
you
can
use
it
as
a
pattern
recognizer,
but
it
also
can
generate
patterns
and
that's
very
important
in
development.
A
So
you
can
run
a
simulation
where
you
have
this
grid.
You
have
these
cells
that
change
their
color
based
on
their
neighbor's
State,
and
you
end
up
with
these
patterns.
So
you
can
get
these
kind
of
dotted
patterns,
you
can
get
stripe
patterns
and
the
like-
and
you
know
he
worked-
he's
worked
with
it
a
little
bit.
He
worked
with
it
with
someone
I
think
in
the
summer
of
2017,
and
it
really
hasn't
seen
any
development,
so
this
might
be
something
to
resurrect
since
this
time.
A
Maybe
some
of
the
methods
like
neural
cas
es
that
might
be
something
for
you,
a
project
for
you
to
do
so.
That's
those
are
medium
public
lectures.
Another
thing
we
have,
of
course,
are
Publications,
and
so
they
have
a
lot
of
Publications
in
the
group.
We've
done
a
lot
of
work,
so
we've
been
working,
we've
been
going
since
2014
and
we
put
out
a
lot
of
papers.
A
We
have
a
lot
of
Publications
and
peer-reviewed
journals
in
pre-print
repositories,
and
you
know
presentations
at
conferences,
large
International
conferences,
smaller
conferences
and
then
we've
of
course
contributed
to
things
like
some
of
the
things
that
openworm
has
done.
We've
contributed
to
a
lot
of
book
chapters.
A
So
all
of
these
you
know
this
is
the
list
of
Publications.
It
looks
pretty
long
which
it
is.
We
have
blog
posts
which
are
posted
on
a
book.
You
know
those
are
rather
informal
but
they're
actually
quite
interesting,
especially
with
respect
to
some
of
the
things
we've
done
in
the
history
of
the
project.
A
C
A
A
A
A
See
it
okay,
all
right
good,
so
some
of
these
posters
are
tied
to
Publications
and
some
are
not.
We've
done
this
one
poster
neuromorphogenetic
patterns
and
the
theory
of
deep
learning
was
kind
of
interesting
because
we
kind
of
tied
and
a
lot
of
stuff
we've
done
with
pattern
formation
to
some
of
the
deep
learning
theory-
that's
out
there,
so
that
was
kind
of
interesting,
but
you
know
we
didn't
develop
a
paper
out
of
it.
A
Some
of
this
is
like
a
new
kind
of
Developmental
biology,
which
is,
you
know
like
basically
describing
what
we're
trying
to
do
here
in
the
group
and
then
other
types
of
you
know:
Project
Specific
posters,
one
of
the
things
about
having
you
know.
We
have
the
poster
online.
So
one
of
the
things
about
having
a
poster
at
a
meeting
is
that
you
see
it
for,
like
maybe.
A
We
also
have
papers,
we've
done
a
lot
of
theory
papers
and
simulation
papers.
We've
done
a
lot,
a
fair
amount
of
comparative
development
where
we've
taken
C,
elegans
and
compared
them
with
different
species.
So
we
have
one
paper
where
we
compared
C,
elegans
and
zebrafish
and
another
where
we
compared
C,
elegans
and
siona
intestinals,
which
is
the
c-squirt
which
is
a
marine
invertebrate
with
a
very
interesting
life
history.
So
that's
or
you
know
where
they
go
through
different
stage,
larval
stages,
so
those
are
interesting
comparative
biology
papers.
A
We've
of
course
done
work
on
the
diatoms
under
this
basilaria
category
drosophila,
which
is
the
fruit
fly.
We've
done
a
couple
papers
on
that
and
then
of
course,
C
elegans,
which
we've
done
papers
and
we've
done.
Data
we've
generated
data
sets
on
that.
So
those
are
all
what
we
have
for
our
Publications
and
so
I
want
you
to.
You
know
if
you're
involved
in
the
Community
period
or
even
if
you're,
not
I,
encourage
you
to
take
a
look
at
those
things.
D
D
D
So
these
are
the
data
points
which
are
more
compressive
again,
that
is
less
compost.
Okay
and
I
tried
to
raise
the
documentation
of
pipelot.
Then
I
was
able
to
understand
what
this
really
this
dark
represent.
So
I
used
I
mean
I
just
converted
a
different
digits
and
I
tried
to
plot
it,
how
it
looks
and
then
I'm
just
seeing
like
whether
it
could
I
try
and
say,
can
I
use
this
data
trying
to
3D
nuclear
segmentation
for
video
program,
so
the
video
is
looking
fine.
These
are
just
these
dots.
D
D
D
B
D
Base,
but
there
is
a
full
business
I
thought
this
would
work
and
I
wasted
around
two
to
three
days
using
this
library,
but
at
the
end
I
came
into
model.
This
code
has
little
parts
so
I'll
put
in
the
user,
then
after
that
I
could
find
two
resources.
Okay,
so
I
was
thinking
to
start
working
on
the
code
as
usual.
D
D
D
D
D
Actually,
for
annotations
part,
we
don't
have
the
ground
truth
actually
like
like
mentioning
that
notation.
This
is
itself
I'm,
just
being
happy
I'll
just
take
that
as
a
cell
I'm,
giving
that
as
only
one
class
was
there,
because
only
once
I
mean
so
I'm
going
to
use
that
we
can
mention
that
in
the
code
and
most
of
the
part
can
be
handled
by
segment
anything
model
because
it
is
so
even
the
parameter
of
our
network
was
696
billion,
which
is
so.
D
D
D
D
C
C
D
I'll
definitely
see
and
I
will
try
to
plot
it
in
3D
structure,
as
you
know,
actually
this
before,
like
difference
and
small
sense,
actually
to
plot
them
in
the
3D
stuff.
Oh,
okay,
okay,
this
would
be
not
like
a
big
problem.
I
will
try
to
create
a
3D
box
around
distance,
so
it
looks
like
a
three-dimensional
in
different
way.
If
you
change
the
viewer,
if
we
move
the
mouse,
the
cells
will
move
a
little
bit
and
we
can
see
that
3d
effect
like
how
this
is.
C
A
D
D
C
C
D
A
Okay,
okay,
that
sounds
good.
Thank
you
for
the
update
yeah.
It
looks
like
some
interesting
stuff
and
looks
like
you
know.
I
know
you're
stuck
on
some
things,
but
that's
that's
part
of
what
we're
doing
in
the
Pro
program.
You
know
trying
to
figure
out
how
to
get
around
some
of
these
bottlenecks
and
if
you
have
like
some
other
things
you
can
do
in
the
meantime,
then
you
know
that
gets
done
and
you
just
kind
of
work
on
that
problem.
So
that's
I
think
that's
a
good
demonstration
of
that.
A
Well,
I
think
you
can
get
an
extension,
but
I
think
you
have
to
wait
until
sort
of
towards
the
end
to
see
if
you
you
know,
can
get
it
extended.
The
only
thing
it
does
is
like
it
just
basically
you
spend
less
time
on
it
per
week
per
week.
You
know
what
I
mean
like
it's,
it's
the
the
project
length
doesn't
change,
it's
just
the
amount
of
time
you
have
to
complete
it.
So
if
you
extend
it
yeah
you
get
fewer
hours
or
you
know,
I.
D
D
She
mentioned
like,
if
you
want
to
increase
your
time,
then
mention
it
to
your,
because
you
need
to
make
changes
in
some
web
app
to
know
them
like
my
project
has
been
extended
according
to
that,
only
funds
will
also
be
released.
My
I
mean
my
first
evaluation.
Timing
is
gonna,
be
changed
from
July
14
to
some
other
periods.
D
A
C
A
I
think
it
still
probably
would
be
like
later
on
in
the
Summer,
where
we
say
that
so
we'll
see:
okay,
okay,
yeah
all
right!
Well,
yeah!
Thank
you
for
that
update
those
very
good
stuff.
It
looks
like
you're
getting
into
a
lot
of
good
toolboxes
and
things
like
that
and
we're
moving
the
ball
on
that
so
and
we're
still
technically
in
the
community
period.
So
that's
good.
We
have
a
couple
more
weeks
before
the
official
coding
period
starts.
It's
good
that
people
are
getting
their
ducks
in
a
row
and
everything.
A
So,
let's
that's
good!
So
now
I'm
going
to
talk
about
some
things
that
I
found
so
I
think
one
of
them
has
actually
from
the
slack
channel
that
we
had.
Another
thing
is
something
that
I
think
will
be
interesting.
We
were
just
talking,
for
example,
about
genetic
data
and
Gene
regulatory
data,
and
last
week
we
talked
about,
like
you
know,
engineering,
mammalian
cells.
Basically,
so
there's
this
field
synthetic
biology
where
people
can
create
they
can
boot
up
cells
to
do
certain
things.
A
Mammalian
cells
are
a
lot
harder
for
a
number
of
reasons,
but
the
the
thing
I
showed
last
week
talks
about
sort
of
the
toolkit
that
we
would
need
to
have
to
build
to
engineer
mammalian
cells,
but
one
of
the
other
things
we
need
is
to
have
methods
to
manipulate
genes,
because
genes,
of
course,
are
the
building
blocks
of
proteins
genes,
get
transcribed
and
oftentimes.
What
happens
in
a
lot
of
the
experiments
you'll
see
in
in
biological
papers
is
where
they
manipulate
a
gene
to
do
something
else.
A
So
their
methods-
and
these
go
back
many
many
decades
where
they
would
mutate
a
gene
or
they
would
manipulate
it
in
some
way.
They
might
knock
it
out.
It
might
knock
out
a
gene
they
might
over
what
they
call
over
Express
the
gene,
which
means
it
gets
expressed
over
what
it
would
be
in.
Normally,
you
can
have
multiple
Gene
copies.
You
can
insert
them
into
the
genome
in
different
places.
You
can
do
all
sorts
of
things
like
that,
but
in
mammalian
cells
and
and
like
in
animal
cells.
A
More
generally,
there's
been
a
lack
of
control.
In
other
words,
you
can
put
the
genes
in
or
knock
them
out,
but
sometimes
you
don't
really
have
good
control
across
cells.
So
you
get
like
you
have
to
do
a
lot
of
selection
of
cells
that
have
that
are
expressing
the
the
gene
in
the
right
place
and
so
forth,
and
so
Gene
editing
is
actually
an
area
that
has
been
sort
of
elusive
but
moving
fast.
A
A
People
have
been
perfecting
over
that
time.
I,
don't
I,
wouldn't
say
it's
perfect,
but
it
it's
it's
a
lot
more
accurate
than
some
of
the
earlier
methods.
So
this
technique
that
I'm
going
to
talk
about
today
is
a
sort
of
a
step
up
from
crispr
and
I'll.
Say
it
right
now.
That
crispr
is
like
this
hard
as
hard
to
wrap
your
head
around
some
of
the
terminology
they
use,
but
basically
they're
selecting
they're
going
down
the
chromosome.
A
This
is
something
called
Prime
editing,
and
so
this
is
like
a
step
up
from
crispr
and
so
Prime
editing
is
actually
like.
I
said
it's
Gene
editing
you
can
go
through
and
you
can
edit
out
genes,
you
can
edit
out
stretches
of
DNA,
you
can
insert
new
stretches
of
DNA,
and
so
in
that
sense
it's
very
highly
controlled.
A
So
so
this
is.
This
is
where
precise
genome
editing
is
very
useful.
So
this
group
from
MIT
the
broad
Institute
St
Jude's
children,
Research
Hospital
they've,
showed
that
a
precise
genome,
editing
approach,
Prime
editing,
can
change
mutated,
hemoglobin
genes
back
to
their
normal
form
and
SCD
patients,
which
restores
normal
blood
parameters
after
transplantation
into
mice.
So
they
do
a
lot
of
things
with
the
cells
you
know
to
to
move
them
from.
You
know,
sort
of
the
the.
B
A
Where
you're
modifying
the
genes
to
the
human,
and
so
they
have
to
do
some
like
they
have
to
do
some
things
with
the
cells
you
know
to
to
get
it
into
the
form
where
you
can
actually
use
it
for
therapy.
So
it's
kind
of
an
interesting
field,
because
you
know
they're
always
trying
to
figure
out
ways
to
do
therapies
and
again
they
don't.
This
isn't
something
that's
approved
by
the
FDA.
A
It's
kind
of
like
these
proof
of
concept
things,
so
this
was
published
in
nature,
biomedical
engineering
and
it's
just
an
example
of
this
Prime
editing
technique.
So
the
prime
editing
technique
is,
let's
see
if
they
talk
about
it
here,
actually,
maybe
not
in
this
paper.
But
this
is
just
the
paper
where
they
show
the
application
to
sickle
cell
and
then
this.
So
the
prime
editing
approach
is
a
promising
approach,
because
in
theory
we
can
directly
correct
disease
mutations
to
specify
healthy
DNA
sequences
of
our
choosing.
A
We
optimize
Prime,
editing
and
long-term
blood
stem
cells
and
show
that
the
prime
editing
cells
maintain
full
engraftment
efficiency
in
an
animal
in
a
clinic
with
a
clinically
relevant
system.
So
you're
able
to
fix
mutations
using
prime
editing,
they're
able
to
take
the
system
and
they
can
find
the
disease-causing
mutation
we
have
Genesee.
We
have
our.
You
know
the
human
genome
sequenced.
A
We
can
sequence
genomes
more
and
more
efficiently,
so
we
know
kind
of
the
position
positional
locations
of
these
mutations,
and
so
we
can
use
these
techniques
in
tandem
with
that
information
to
fix
a
specific
mutations.
So
this
is
this
is
all
coming
together
to
sort
of
treat
diseases,
so
you
can
replace.
You
know
these.
These
mutations
with
high
efficiency
crime
editing
successfully
corrects
this
mutation
with
up
to
41
conversion
in
blood
stem
cells
from
patients.
A
Previous
research
has
shown
that
editing
and
over
20
percent
of
cells
likely
translates
to
the
therapeutic
benefit.
So
you
know
if
you
go
through
all
the
red
blood
cells,
and
you
just
make
these
edits
to
the
cells-
and
this
is
like
you
know
what
would
be
a
high
throughput
procedure.
You
know
you
could
actually
treat
and
cure
the
disease,
but
you
have
to
have
like
a
system
in
place
where
you
could
cycle
through
the
blood.
It's
almost
like
filtering
the
blood.
A
That
is
because
there's
some
problems
with
that
approach
that
have
to
do
with
the
stability
of
the
DNA
I
believe,
but
the
collaborators
are
previously
shown
base
editing
an
alternative
genome.
Editing
technology
can
turn
mutations
into
a
benign
variant,
but
does
that
doesn't
happen
in
the
original
healthy
sequence?
So
this
is
in
a
nature
publication
where
they
talked
about
this,
where
they're
able
to
treat
this
in
mice-
and
you
know
they
they're-
actually
looking
at
the
function
of
the
blood
cells,
with
respect
to
the
mutation
and
so
yeah.
A
Is
just
heads
up
on
a
new
type
of
technique
that
is
being
developed,
so
this
is
prime
editing,
it's
sort
of
an
advance
on
crispr
cas9,
which
you
we
may
run
into
in
the
literature,
and
so
that's
a
that's.
You
know
one
of
these
Cutting
Edge
techniques,
so
hi
Morgan.
How
are
you
I
think
speaking
of
Cutting
Edge
techniques?
A
That's
what
Morgan
likes
to
hear
about
all
right!
So
that's!
That's
all
I
have
to
talk
about
for
that,
and
then
this
other
paper
that
came
up
in
the
slack.
It's
I
think
in
the
diva
learned,
Devo
graph,
Channel
or
I'm,
not
really
sure
where
it
is
it's
in
one
of
the
channels
either
that
or
the
devil
worm
Channel.
B
A
Is
going
back
to
the
cell
engineering
idea
where
we
want
to
like
engineer
cells
and
want
to
improve
their
function?
We
want
to
change
their
function
and
to
do
that,
though,
in
mammalian
cells
say,
we
need
to
know
a
lot
more
about
how
cells
work
and
we
don't
fully
know
how
cells
work,
but
we
can
use
some
of
the
principles
that
we've
learned
from
say
systems
biology
to
do
this,
and
so
one
of
the
things
we've
learned
is
that
cells
have
this
aspect
of
modularity.
A
So
you
know
this
is
something
that
we
can
use
to
program
these
complex
functions
last
week,
I
talked
about
the
different
ways.
We
can
do
that,
but
this
week
I
want
to
talk
about
this
idea
of
modularity,
so
a
New
York
biological
engineering
is
emerging
which
living
cells
are
used
as
building
blocks
to
address
therapeutic
challenges.
A
These
efforts
are
distinct
from
traditional
molecular
engineering.
Their
focus
is
not
at
optimizing
individual
genes
in
proteins.
So
this
is
what
we
talked
about
in
the
case
of
prime
editing,
but
rather
on
using
molecular
components
as
modules
to
reprogram
how
cells
make
decisions
and
communicate
to
achieve
higher
order.
Physiological
functions,
so
this
is
the
cell-centric
approach,
as
opposed
to
the
gene,
Centric
approach
and.
A
Think
modifying
genes
is
good
for
treating
diseases
and
targeting
specific
functions
where
you
know
what
the
underlying
genetics
are.
If
you
don't
know
that,
then
the
self-centric
approach
makes
more
sense,
and
so
the
cell-centric
approach
as
a
growing
toolkit
of
components
that
can
synthetically
control,
core
cell
level,
functional
outputs
such
as
where
in
the
body
a
cell
should
go.
A
You
know
this
is
the
stuff
that
Michael
Levin
does
in
some
of
his
work
with
flatworms,
trying
to
think
about
like
the
whole
organism
and
sort
of
how
cells
can
you
know,
regenerate
and
find
a
place
in
the
organism,
and
so
but
people
have
been
thinking
about
this
for
a
long
time.
This
is
a
very
common
problem
in
developmental
biology,
for
example,
but
we
we
can
do
this.
You
know
either
by
sort
of
letting
nature
do
its
thing
and
then
assisting
nature,
or
we
can
actually
Design
Systems.
A
That
would
allow
that
that
sort
of
where
we
can
control
this,
so
maybe
we
can
put
cells
in
places
where
they
wouldn't
normally
go.
So,
if
you're
doing
stem
cell
therapy,
for
example,
you
have
a
stem
cell
that
you're
differentiating-
and
it
normally
may
not
know
where
to
go-
or
you
know,
go
in
the
right
place
because
it
may
derive
from
say
you
know
some
other
type
of
tissue.
Well,
you,
if
you
know
the
rules
of
the
system,
you
know
how
to
guide
those
cells
into
the
right
place.
A
So
this
is
the
type
of
thing
we're
wanting
to
do
here.
So
talk
about.
You
know
how
this
can
be
used
by
forcing
the
conceptual
distillation
of
complex
biological
functions
into
a
finite
set
of
instructions
that
operate
at
the
cell
level.
These
efforts
will
also
ship
light
on
the
fundamental
hierarchical
Logic,
the
Link's
molecular
components
to
higher
order
physiological
function.
A
So
they
talk
about
modularity
here
they
talk
about
the
sort
of
set
of
instructions,
so
this
is
a
graphic
here
where
you
have
a
change
in
the
genome
in
molecules,
and
so
this
is
the
DNA
here
this.
These
are
the
like
membrane,
proteins
and
other
types
of
proteins
that
you
know
get
formed
that
do
things
in
the
cell.
Then
you
have
things
at
the
cellular
level,
where
you
get
inputs
from
other
cells.
A
There's
this
decision-making
circuit
when
they
say
decision
making,
they
don't
mean,
like
the
cell,
thinks
they
mean
that
it
either
does
one
thing
or
another,
and
there
are
all
sorts
of
different
genetic
circuits
that
allow
it
to
make
these.
So
in
other
words,
the
cell
can
decide
to
live
or
die,
and
so
the
cell
can
either
go
apoptose
or
not
that's
a
decision-making
thing,
and
so
there's
a
change
there
and
it's
triggered
by
some
input
from
other
cells.
It's
something
that's
done
by
through
genetic
regulatory
circuits
and
then
that's
the
output
to
other
cells.
A
A
So
then,
you
have
the
solid
state
toolkit
that
that
allows
you
to
manipulate
these
cells.
So
they
can
do
these
things.
You
have
a
toolkit
that
allows
you
to
transmit
signals.
Sun
signals
move
and
change
its
shape,
so
the
cell
can
change
its
shape.
We've
talked
about
that
a
lot
but
to
have
control
over.
That
is
the
thing
that
they're
proposing
here:
death
cell
death,
the
cell
dies
differentiation,
proliferation
and
adhering
to
a
surface.
So
the
idea
is
you
have
these
modules
they
think
of
this
as
modules.
A
A
So
we
have
this
toolkit,
and
this
is
again
I
think
this
is
more
kind
of
a
aspirational
thing,
so
they
don't
propose
an
actual
set
of
tools,
but
they
say
that
you
have
this
tool
kit
and
ideally
it's
like
a
artist
palette
where
you
have
different
colors
color
gradations,
and
you
pick
one
and
you
put
it
on.
You
know
you
turn
it
on
and
you
see
if
it
works,
you
try
another
one
to
see
if
it
works,
and
then
you
move
to
another
function
and
you
turn
it
on.
A
So
yeah
we
want
to
have
this
Universal
toolkit
and
the
whole
idea
is
basically
being
able
to
push
buttons
and
get
outputs
that
we
can.
You
know
basically
control
like
a
button
press
which
maybe
again
like
in
biology.
Maybe
that's
a
wrong
headed.
Maybe
that's
fundamentally
misguided,
but
that's
the
goal
here.
So
you
know
we
have
these.
We
actually
know
that
there
are,
you
know,
genetic
regulatory
networks
for
certain
functions.
We
know
like
in
some
diseases
where
we
have
specific
genetic
mutations
that
turn
diseases
on
or
off
or
modify
them
some
way.
A
So
we
have,
we
know
those
exist,
but
we
don't
really
have
a
tool
kit
for
that,
and
sometimes
we,
you
know
for
some
diseases,
it's
a
little
bit
easier
than
say
other
things
like
a
change
in
shape.
Could
we
say
that
you
know
if
you
just
change
these
mutations
or
if
you
knock
this
Gene
out
and
it
changes
shape
I
mean
you
know.
There
are
a
lot
of
the
papers
where
they
do
these
kind
of
experiments.
But
the
question
is:
is
how
accurate
do
you
get
I
mean
you
know?
Can
you
really
control
that?
A
And
so
that's
not
the
idea,
and
then
from
that
that
toolkit
you
can
then
go
and
have
like
customize
cell
response,
so
we
might
be
able
to
say,
recognize
disease
tissues
and
when
you
see
a
cell,
that's
diseased
versus
not
diseased.
You
can
destroy
that
cell
turn,
it
off
and
say
die
versus
the
healthy
cells.
A
You
can
say
you
need
to
live
and
you
can
push
that
button
and
the
idea
would
be
that
you'd
be
able
to
replace,
say
cancer
cells
with
normal
cells
and
so
or
or
favor
normal
cells
over
cancer
cells
and
needles
have
all
engineered
multicellular
networks.
So
you
have
these
kind
of
networked
interactions
between
cells
that
you
can
also
control,
and
so
this
is
yes
kind
of
goes
through
some
of
these
things,
where
you
have
these
cellular
tasks
like
delivering
suppressive,
payloads,
identifying,
inflammatory
sites
identifying
and
killing
cancer
cells.
A
So
there
are
all
sorts
of
things
that
you
might
want
to
do
and
then
they
have
this
I.
Think
all
the
cell
chassis,
which
are
at
this
point,
I,
guess
just
kind
of
lists
of
cells
that
you
could
use
so
NK
cells,
T
cells.
We
have
a
lot
of
different
cell
types
that
we've
developed
in
the
lab,
and
so
those
you
know
are
useful
for
kind
of
driving
this
forward,
so
you
wouldn't
use
like
normal
cells.
A
Some
of
these,
like
NK
cells,
are
very
specialized
cells,
so
you
know
these
are
this
is
this
is
moving
towards
a
toolkit
I
think,
and
they
show
that
they're
all
sorts
of
different
potential
therapeutic
areas
of
application,
from
synthetic
development
of
organs
to
regenerating
systems
to
reducing
neural
inflammation,
in
reducing
their
degeneration,
reducing
inflammation
and
destroying
cancer,
so
that
these
are
all
you
know,
novel
goals,
but
I,
don't
know
if
we're
really
kind
of
anywhere
near
this.
So
just
to
keep
that
in
mind.
A
So
this
quote
says:
engineering
higher
order,
functions,
forces
us
to
Grapple
with
the
underlying
logic
and
hierarchy
of
biology,
and
indeed
there's
this
like.
We
know
a
lot
about
sort
of
that.
We
know
that
bio
biology
is
hierarchical.
We
understand
there's
some
logic,
especially
in
terms
of
genetic
circuits,
but
we
don't
really
know
the
intermediate
area
because
it
all
fits
together.
So
is
that
something
we
learned
through
engineering
or
through
like
Theory
and
experiment,
it's
hard
to
say.
A
A
So
the
first
paper
is
from
E-Life.
It
was
published
in
2021,
and
the
title
of
this
paper
is
control
of
tissue
development
and
cell
diversity
by
cell
cycle
dependent,
transcriptional
filtering.
So
the
author
has
introduced
this
term
transcriptional
filtering
so
the
abstract
reads:
cell
cycle
duration,
changes
dramatically
during
development
starting
out
fast
to
generate
cells
quickly
and
then
slowing
down
over
time,
Z
organism
matures.
A
A
A
Cell
division
starts
to
slow
down.
Then
you
get
these
long,
Gene
transcripts,
transcribed
in
full,
so
basically
you're
not
getting
all
the
information
you
need
early
on,
so
that
might
impact
function,
so
this
actually
acts
as
a
filter.
So
so
we
generally
think
of
cell
proliferation.
Is
this
thing
that
controls
the
number
of
cells
and
a
tissue,
or
maybe
the
size
of
a
tissue?
But
there's
also
the
secondary
aspect
where
it's
actually
producing
transcripts
differentially?
So
these
mathematical
simulations
to
sort
of
Identify
some
of
these
things,
the
act
it
acts
as
a
tuning
knob.
A
A
Our
predictions
are
supported
by
comparison
to
single
cell
rna-seq
data
captured
over
embryonic
development,
so
they're
using
rna-seq
data
to
sort
of
confirm
this.
So
this
is
where
you
have
sequencing
the
transcripts
and
looking
at
their
sequences
and
seeing
how
much
is
is
present
or
absent
in
a
certain
cell.
A
Additionally,
evolutionary
genomic
analysis
shows
that
fast
developing
organisms,
which
means
organisms
that
have
a
shorter
development
where
they
have
a
quicker
Pace
to
certain
tissues.
If
you
know
the
the
time
it
takes
to
build,
say
like
a
wing
or
some
other
structure
shrinks,
which
happens
in
development,
then
this
can
affect
this
as
well.
A
Evolutionary
genomic
analysis
shows
that
fast
developing
organisms
have
a
narrow,
genomic
distribution
of
Gene
lengths.
So
if
it's
a
fast
development
organism-
and
it
has
these
short
cell
cycles,
the
gene
lengths
are
very
sort
of
concentrated
into
a
certain
set
of
lengths,
and
so
that
means
a
certain
set
of
transcripts
are
being
functionally
expressed.
A
Well,
slower
developers,
meaning
that
they
go
from
this
short
Soul
cycle
to
a
longer
cell
cycle,
because
they
have
a
longer
pace
of
development
as
an
expanded
number
of
long
genes.
So
that
means
that
longer
genes
are
expressed
in
these
organisms
that
longer
transcripts
are
fully
or
longer.
Genes
are
fully
expressed
into
complete
transcripts
and
that
can
have
an
effect
on
function.
A
A
A
If
there's
a
longer
developmental
period,
the
brain
grows
to
a
larger
size.
If
growth
has
accelerated,
it
grows
to
a
larger
size,
sometimes
structures
within
the
brain,
sometimes
the
brain
itself,
and
so
you'll
see
this
across
the
Tree
of
Life
in
different
ways,
and
what
this
suggests
is
that
this
had
these
heterochronic
parameters
have
sort
of
an
effect
on
the
on
the
tissue
and
ultimately
affecting
cell
cycle,
but
that
they're
transcriptional
consequences
to
this,
and
maybe
those
are
important
in
selecting
those
parameters
across
species.
A
Our
results
support
the
idea
that
cell
cycle
Dynamics
may
be
an
important
across
multicellular
animals
for
controlling
genix,
transcript
expression
and
cell
fate.
So
this
is,
you
know,
definitely
they're
proposing
this
mechanism
and
I
think
it's
an
interesting
proposal,
because
I
think
it's
probably
going
to
have
some
sort
of
an
effect.
What
the
actual
effect
is
on
growth
and
development-
it's
not
quite
clear,
but
we
can
see
evidence
of
it
in
the
in
the
data.
A
So
a
fundamental
question
in
biology
is
how
a
single
cell,
eukaryotic
cell,
or
a
single
eukaryotic
cell
zygote
stem
cell,
produces
the
complexity
required
to
develop
into
an
organism.
A
single
cell
will
divide
and
generate
many
progeny,
which
means
they
generate
a
lot
of
different
Offspring
cells
or
daughter
cells,
diversifying
in
a
controlled
and
timely
manner.
A
So
you
can
see
that
there's
these
Force,
these
four
phases
of
cell
division
cell
cycle,
so
it
goes
from
dividing
to
replicating
its
DNA
and
then
dividing
again,
and
so
the
idea
here
is
that
there
there's
a
control
on
these
phases
and
there's
a
whole
genetic
circuit.
That's
involved
in
this
where
it
regulates
the
timing
of
it.
So
if
you
shorten
the
first
Gap
phase,
you
can
shorten
cell
cycle.
If
you
shorten
the
synthesis
phase,
you
can
also
shorten
the
cell
cycle.
If
you
lengthen
this
synthesis
phase,
you
lengthen
cell
cycle.
A
One
of
the
consequences
of
that,
though,
is
that
you
lengthen
the
time
in
which
these
things
can
take
place.
So
you
might
have
things
going
on
during
that
time.
That
are,
you
know,
have
consequences.
So
first,
you
know
in
law
an
elongated
synthetic
phase
that
not
only
has
an
effect
on
the
entirety
of
cell
cycle,
but
also
it
has
an
effect
on
what's
going
on
during
that
phase
in
the
cell's
metabolism,
so
the
length
of
each
phase
determines
how
much
time
a
cell
allocates
for
the
process
associated
with
growth
and
division.
A
This
is
true,
I
think
for
model
stem
cells,
but
also
for
differentiated
cells,
and
you
can
definitely
have
you
know,
a
change
in
the
cell
cycle
for
depending
on
the
function
of
the
cell,
so
this
is
kind
of
what
led
them
to
the
study
to
think
about
like
okay.
Why
is
it
that
you
have
this
change
across
cell
types.
A
Organisms
such
as
the
fruit
fly
and
the
worm
drosophone
C
elegans,
respectively
exhibits
cell
cycle
durations,
as
short
as
eight
to
ten
minutes
cell
cycle
duration
also
changes
over
development.
For
example,
it
increases
in
Mouse
brain
development
and
from
an
average
of
eight
hours
to
at
embryonic
day,
11
to
an
average
of
18
hours
by
organic
day,
17..
So
there's
a
shortening
of
time.
There's
this
difference
in
time
across
cell
types,
but
there's
also
this
difference
in
time
across
the
organisms.
A
A
B
A
It
does
have
a
number
of
consequences
in
probably
we're
not
talking
about
symptoms
so
interesting.
Interestingly,
cell
cycle
duration
can
act
as
a
transcriptional
filter
that
constrains
transcription.
In
particular,
if
the
cell
cycle
progresses
relatively
fast,
transcription
of
long
genes
will
be
interrupted
and
then
typical
cells
of
Gene
transcription
rate
is
between
1.4
and
3.6
kilobases.
B
A
Minute
so
that
means
it
goes
down
the
DNA
strand
and
it
transcribes
like
something
like
1400
to
3600
bases
per
minute
or
I.
Guess
it
would
be
yeah,
so
it
would
be
well
I,
guess
yeah.
That
would
be
right.
So
this
is
a
pretty
fast
moving
polymerase,
so
you're
getting
a
lot
of
things
transcribed
per
minute.
If
you
think
about
it
that
great
doesn't
necessarily
change
you're,
just
transcribing
more
or
less
so.
If
your
cell's,
like
a
shorter,
you
get
less
transcript.
A
Now,
that's
a
pretty
large
range.
So
that's
you
know.
We
don't
really
know
how
that
varies.
Maybe
it
varies
a
cell
cycle,
so
you
get
some
compensation
for
a
shorter
cell
cycle.
So
in
other
words
it
goes
faster
with
a
shorter
cell
cycle.
It
may
be
slower
with
longer
cell
cycle,
but
I,
don't
I,
didn't
I
haven't
looked
at
that
reference,
so
maybe
that's
not
the
case.
B
A
Just
shows
you
cell
cycle
duration
changes
during
Mouse
development,
so
this
is
the
different
cell
types
in
Mouse
from
embryo
to
trophoblast
and
mesoderm
to
retina
neural
tube
cerebral
wall
cortical
stem
cell-
this
is
just
a
large
actually
retinas
over
here.
So
this
is
a
large,
very
a
large
variety
of
different
cell
types,
and
this
is
the
number
of
days
in
which
you
have
some.
A
This
is
the
cell
cycle
duration.
Here
in
hours
on
the
y-axis
on
the
x-axis,
you
have
days
of
development,
so
you
see
that
the
embryo
is
pretty
early
in
development,
embryo
cells.
You
have
the
different
germ
layers
here.
You
have
Retina
cells
early
in
retina
cells,
late
retinas
cells
in
the
middle,
and
you
can
see
that
the
retina
cell,
if
we
look
at
the
retina
song,
We
track
it
across
time
at
seven
days
of
development.
A
The
cell
cycle
duration
is
very
short
and
it
increases
from
like
a
Poe
I
guess
about
four
hours
to
about
11
hours
from
day,
seven
to
day
10.
and
then
to
day
15
it's
about
20
hours
of
duration
and
then
actually
at
day,
22
it's
about
30
hours
generation.
So
you
can
see
that
they're
actually
Four
data
points
here,
that's
a
pretty
good
trend
of
like
over
developmental
time
and
increasing
cortical
stem
cells,
on
the
other
hand,
are
less
it's
less
clear.
A
You
have
a
bunch
here
at
about
the
same
time
in
development
and
they're
kind
of
varying
in
the
number
of
hours
for
their
cell
cycle
duration.
That's
actually
probably
not
surprising,
since
the
stem
cells
are
very
potent,
and
so
they
can
form
a
number
of
different
cell
types,
whereas
retin
has
already
differentiated.
In
any
case,
this
is
what
it
looks
like
you
get
this
tendency
towards,
as
you
get
further
along
in
development
of
higher
cell
cycle
durations
or
larger
than
a
number
of
hours.
A
Going
on
there
again,
they
may
it
may
be
because
they're,
pluripotent
or
even
tonypope-
we
don't
really
know,
but
in
any
case,
so
we
asked
what
effects
cell
cycle
dependent,
transcriptional
filtering
may
have
over
early
multi-cellular
organism
development,
their
extensive
mathematical
simulations
of
Developmental
cell
lineages.
We
identify
the
novel
and
unexpected
finding
with
a
cell
cycle,
dependent
transcriptional
filter
and
directly
influence
the
generation
of
cell
diversity
and
can
provide
fine-grained
control
of
cell
numbers
and
cell
type
ratios
in
a
developing
tissue.
A
This
is
not
a
conscious
thing,
but
this
is
something
that
happens
just
through
this
switch,
so
there's
a
switch
that
happens
or
some
sort
of
gradient
gradation,
but
then
I
was
for
this
kind
of
control,
and
so
again
you
know
we
have
this.
So
this
is
cell
cycle
duration.
This
is
a
cell
with
a
set
of
transcripts
being
produced.
A
The
gene
length
is
the
introns
and
exons,
so
it's
like
the
coding
regions
and
the
non-coding
regions
all
together,
and
this
is
actually
an
interesting
point,
because
the
longer
the
transcripts,
the
more
things
that
are
being
produced.
So
in
other
words,
you
know,
you'll
have
a
lot
of
opportunities
for
alternative
displacings,
for
example,
where
the
RNA
is
spliced
in
different
ways
to
take
different
parts
of
different
introns
from
that
Gene
and
use
it
for
different
things.
So.
A
Sort
of
enhanced
regulatory
capacity
when
you
transcribe
shorter
Gene
lengths
you're
only
getting
part
of
the
gene,
perhaps
you're
only
getting
the
sort
of
the
main
function
of
the
gene.
It's
not
really
clear
to
me
and
I'm,
not
an
expert
at
late,
the
structural
biology
of
genes,
but
what
that
consequence
is
or
if
it's
even
something,
that's
predictable,
but
the
idea
that
you're
putting
out
introns
and
exons
means
that
you
have
this
opportunity
for
alternative
slicings
or
alternative
isofore
isoforms
of
different
genes,
and
so
this
allows
you
to
have
an
enhanced
function.
A
So
this
shows
that
the
rna-seq
so
you're
actually
looking
at
some
of
these,
you
have
different
isoforms
and
you
have
different
genes
and
you
can
actually
get
the
transcript
counts.
This
is
basically
what
they're
doing
in
mRNA
seek
so
they're,
actually
looking
at
the
transcript
count
for
Gene,
one
for
Gene
lion
and
team,
two
and
Gene
one
gene
2
and
Gene
3..
A
A
If
you
lengthen
the
duration
of
cell
cycle
a
bit
more,
you
get
the
second
Gene
that
pops
up,
but
you
also
get
two
isoforms
of
Gene
one,
and
then
you
can
see
that
here
and
then
you
not
only
get
if
you
lengthen
the
cell
cycle
even
further.
You
not
only
get
all
three
of
these
genes
in
the
circuit,
but
you
also
get
three
isof
forms
of
one
and
it
could
also
be
three
copies
of
the
same
transcriptions.
You
know
this
is
the
thing
about
like
lengthening
the
window
of
transcription.
A
A
This
happens
to
affect
cell
division,
because
when
you
have
a
longer
cell
cycle,
you
get
more
transcripts
that
can
be
randomly
distributed
to
the
daughter
cells
and
that
leads
to
different
types
of
transcripts
being
transmitted
from
other
cell
to
daughter
cell,
and
it
leads
to
differential,
like
differentiation
in
in
different
ways.
So
that's
basically
what
their
argument
is
and
I
think
it's
an
interesting
paper,
because
it
leads
to
a
lot
of
questions
about
what
the
function
of
cell
cycle
is
and
what
is
sort
of
being.
A
A
You
model
the
cell
cycle
as
certain
lengths
and
then
you
see
which
ones
are
possible
within
that
interval.
So
it's
very,
very
simple
simulation
numeric
simulation,
and
so
this
is
the
output
here
where
the
filter
is
off
versus
when
the
filter
is
on.
But
actually
when
the
filter
is
on,
you
get
more
clusters,
at
least
on.
The
cell
cycle
was
around
five
in
duration
versus
later
or
earlier,
and
then
so.
Transcriptome
diversity
actually
is
affected
by
being
off
versus
clusters
being
affected
when
it's
on,
and
then
this
shows
like
there's
a
t.
A
A
What
they
mean
by
that
is
so
cells
are
the
longer
cell
cycle
duration,
the
blue
lineage,
generates
viewer
progeny
with
respect
to
the
cells
or
the
short
cell
cycle
duration
of
one
hour,
which
is
the
gray
lineage,
and
so
the
blue
again
is
the
cell
cycle
duration
and
the
gray
is
the
cells
of
the
short
cell
cycle
of
duration.
So
these
longer
cell
cycle
durations
generate
fewer
progeny.
A
The
slower
cells
contribute
more
to
the
diversity
observed
in
the
population,
so
increasing
cell
cycle
duration
increases
cell
diversity,
but
also
limits
the
number
of
progeny
generated.
So
we
see
that
in
fast
fast
we
have
two
sub
lineages
that
form
the
blue
and
the
gray
they're
they're
symmetrical.
A
When
we
have
a
slow,
fast
scenario,
when
we
slow
down
the
division
rate
of
the
blue,
we
only
have
four
Cells
versus
eight
cells.
At
the
end
of
the
simulation,
so
for
the
fast
fast
scenario,
transcriptional
diversity
is
really
low.
It
increases
with
the
slow
fast
scenario
and
then
the
proportion
of
blue
cells
and
the
cell
percentage
here.
A
The
cells
that
survive
this
division
process
is
almost
increa.
Incompletely
blue
in
the
transcriptional
diversity
is
high,
whereas
in
slow
fast
in
this
case
the
gray
cells
dominate
the
transcriptional.
Diversity
is
also
high.
In
this
case,
the
cell
percentage
is
equal
in
the
fast
fast,
but
the
transcriptional
diversity
is
low.
A
Okay,
so
that's
that
paper.
A
second
paper
is
this
Coral
Court
control
principles
of
the
eukaryotic
cell
cycle.
So
now
we
go
from
this
idea
of
transcriptional
filtering
this
idea
of
control
principles
for
the
cell
cycle
itself.
So
that
makes
sense,
because
we
want
to
know
what's
underlying
this
phenomena
and
it
involves
controlling
cell
cycle.
A
However,
the
principle
on
which
Cyclone
cdk
complex
is
organized
the
temporal
order
of
cell
cycle
events
or
contentious.
So
the
two
competing
models
here
and
we
don't
really
know
how
these
things
are
ordered
temporarily
and
how
they're
controlled
the
first
model
is.
This
scdks
and
mcdks
are
functionally
specialized,
so
SD
cdks
are
these
different
Cyclone
CD
for
DNA
replication.
A
A
second
model
proposes
that
scdk's
mcdk
is
redundant
with
each
other,
both
acting
as
sources
of
overall
cdk
activity.
In
this
model,
increasing
cdk
activity,
rather
than
cdk
substrate
specificity
or
our
cell
cycle
coordinates.
So
it's
actually
activity
or
a
specificity
in
the
second
model.
In
the
first
model.
It's
the
idea
that
you
need
specificity.
A
Here.
We
reconcile
these
two
views
of
poor
cell
cycle
using
us
for
phosphoridomic
assays
of
in
Vivo
cdk
activity
and
vision.
East,
we
find
the
cscdk
and
mcdk
substrate
specificity
is
a
remarkably
similar
showing
that
scdks
and
mcdks
are
not
completely
specialized
for
S
phase
of
mitosis
alone.
So
these
are
the
the
DNA
replication
S
phase
and
my
division
cell
division
of
mitosis.
A
A
Normally
scdk
cannot
drive
mitosis,
but
can
do
so
in
protein.
Phosphatase
one
is
removed
from
the
centrism,
thus
increasing
your
cdk
activity
in
Vivo
is
sufficient
to
overcome
substrate
specificity
differences
between
scdk
and
mcdk,
and
allows
us
cdk
to
carry
our
mcdk
function.
So
you
have
this
interchangeability,
and
there
is
a
lot
of
you
know.
Increasing
activity
can
overcome
substrate
specificity
in
certain
conditions,
and
so
we
have
this
sort
of
I
guess
you
could
say
biological
adaptability
or
even
biological
robustness.
A
Therefore,
we
reunite
the
two
opposing
views
of
cell
cycle
control,
showing
that
the
core
cell
cycle
engine
is
largely
based
on
quantitative
increases
in
cdk
activity
through
the
cell
cycle.
So
we
have
these
increases
in
cdk
activity
and
that's
what
drives
different
parts
of
the
cell
cycle
on
its
timing,
combined
with
minor
and
ceremonial
qualitative
differences
in
catalytic
specialization.
A
A
So
this
kind
of
goes
through
some
of
their
data.
It's
a
pretty
deep
paper
in
terms
of
technical
specificity,
so
for
this
group
I,
don't
think
that
you're
willing
to
get
a
lot
out
of
it
by
going
through
the
details,
I
think
that
this
does
kind
of
complement
this
other
paper
and
then
it
gives
a
mechanism
for
what's
going
on
there.
A
The
other
paper
is
just
kind
of
like
looking
at
the
MRNA
and
seeing
you
know
what
is
the
expression
model
hypothesize,
that
shorter
timing
will
lead
to
Shorter
amounts
of
transcription,
and
we
can
confirm
that
with
a
data
set.
This
shows
you
the
actual
mechanisms
behind
it
and
some
of
the
biochemistry.
So
this
is
an
interesting
set
of
papers
and
I
think
it
has
a
lot
of
consequences
for
thinking
about
cell
division
and
some
of
the
variation
we
see
across
species
and
across
developmental
time.
A
A
A
We
will
go
over
those
and
replace
the
slides
where
you
didn't
achieve
what
you
wanted
to,
or
you
achieved
much
more
than
you
wanted
to,
or
that
you
actually
did
some
work
on
it
and
then
you'll
present
that
again-
and
the
point
is-
is
to
show
the
progress
and
sometimes
the
detours
that
you
have
to
make
so
I,
don't
know
when
we'll
present-
those-
probably
not
maybe
not
next
week
but
the
week
after
and
that'll
be
in
the
coding
period.
But
you
know
it'll
be
fine
and
then
it
looks
like
social
month
is.