►
Description
Open project meeting for Open Source Antibiotics Series 1, the Mur Ligases.
Full Project: https://github.com/opensourceantibiot...
Relevant GitHub Issue: https://github.com/opensourceantibiotics/murligase/issues/100
On the call: Prof. Matthew Todd, Yuhang Wang, Yiwei Wang (UCL), Prof Chris Dowson (chair), Dr Adrian Lloyd and Dr Laura Diaz Saez (University of Warwick), Dr Joe Eyermann, Dr Lori Ferrins (Northeastern University), Bart Staker (SSGCID), Jan Abendroth (UCB Pharma).
A
Just
wanted
to
start
that
off
and
it's
the
11th
of
July
open
source
antibiotics,
my
liga's
meeting
so
yeah.
You
recently
Adrian
sent
through
a
file
of
recent
ic50
determination
against
the
enamine
compounds.
If
you're
happy
to
talk
about
that,
that
would
be
really
great.
Okay,.
B
I
know
right.
B
Great,
so,
by
way
of
quick,
brief
introduction,
we
screened
around
the
main
library
and
it's
got
a
very
long
story.
Short.
We
identified
11
extra
members
of
that
Library,
which
appeared
to
be
Murdy
Inhibitors.
B
They
were
characterized
at
single
concentrations
of
compound,
so
there's
no
way
of
knowing
exactly
what
the
potency
is
or
they're
or
if
they
have
any
unusual
behavior
with
respect
to
the
compounds
themselves.
So
we've
started
doing
ic50s,
we've
completed
the
data
collection
and
on
those
11
compounds
and
I'm
working
my
way
through
Computing
the
data,
and
so
these
These
are
the
first
five.
B
B
Things
you
don't
want
to
see,
so
this
is
an
ic50
of
a
molecule
called
k16
from
one
of
the
enemy
plates.
It
has
a
nice
C50
of
167
micromolar.
But
if
you
look
at
this
shape
of
the
profile,
it's
extremely
sigmoid,
you
can
measure
the
sigmoid
density
by
virtue
of
the
hilco
emissions,
where
perfect
hyperbola
has
a
nail
coefficient
of
one
negative
cooperativity
less
than
one
positive
cooperativity
more
than
and
the
sigma
and
nature
the
curve
and
the
hill
coefficient
suggests
marked
positive
cooperativity
and
usually
people
interpret
this
as
aggregation.
B
B
No,
it
can
tell
you,
under
certain
circumstances
how
many
molecules
are
actually
involved.
B
So,
for
example,
the
classic
case
is
hemoglobin
and
oxygen
and
the
whole
coefficient
of
four,
but
in
this
particular
case
no,
no
I'm,
afraid
not.
Okay,
thanks
another
stigmoid
inhibitor
M23,
with
an
ic50
of
just
under
100
micromolar,
and
now
things
get
a
little
bit
better.
So
this
is
compound
m17
from
one
of
the
another
one
of
the
enemy
plates.
There's
a
nice
hyperbolic
relationship
between
inhibition
and
compound
concentration.
B
Hill
coefficient
is
1.3
plus
and
minus
0.3,
so
it
probably
is
close
to
one
and
if
you
were
looking
for
something
that
probably
has
a
single
type
of
interaction
and
then
you
might
want
to
take
forward
that.
B
Similarly,
compound
mo8s,
this
has
a
much
more
attractive.
Ic50
I
have
25
micromolar.
B
That,
basically,
is
in
the
region
of
what
we
were
seeing
with
j06.
Previously
it
means
it
actually
binds
tighter
than
the
standard
inhibitor
we
use
in
our
assays,
which
is
adbcp
and
again.
That
will
be
something
that
I
would
earmark
if
I
wanted
to
take
things
forward,.
B
And
finally,
compound
a19
he'll
coefficient
on
less
than
one
which
in
principle
would
suggest
it's
binding
in
under
a
mode
of
negative
cooperativity,
which
means
binding
of
one
molecule
this
ways
binding
of
another.
Now,
given
the
accuracy,
the
data
that
may
have
been
up
of
the
case,
it
has
a
fairly
High
ic50
I,
wouldn't
eliminate
it
in
consideration,
but
it
may
not
be
the
top
of
my
list,
so
we've
got
another
six
compounds
to
chew
our
way
through
which
I
should
have
done
by
the
end
of
this
week.
B
I
apologize
not
having
them
done
today,
but
we'll
have
the
data
soon
and
that's
kind
of
where
we
are
at
the
moment.
The
plan
really
is
to
then
take
the
compounds,
the
attractive,
looking
ones
and
then
progress
them
through
the
other
ligases
and
see
where
we
are
oh
and
also
one
other
thing
we
have
to
do.
Is
we
have
to
repeat
these
measurements
in
different
ATP
concentrations,
to
determine
the
impact
of
ATP
concentration
in
ic50,
because
what
we
really
really
want
is
an
ATP
competitive,
inhibitor
right.
A
And
there
we
are
that's
wonderful.
Thank
you
just
a
quick
question,
so
solubility
of
these
compounds
has
not
been
a
problem.
B
Well,
if
I
go
back,
I
think
yeah.
So
actually
the
compound
in
front
of
you,
the
crap
in
front
of
you
now
that
will
never
actually
reach
100
inhibition.
B
B
B
So,
quite
often,
from
a
practical
point
of
view,
you're
sort
of
looking
at
the
data
and
having
to
chop
off
large
amounts
of
the
top
end,
simply
because
you're
just
getting
noise
simply
because
you're
getting
so
much
light.
Scattering
all
the
data
I've
shown
you
they.
B
Those
compounds
do
appear
to
have
the
solubility
required
for
me
to
do
the
interpretation
apart
from
this
last
one,
where
I've
only
extended
the
data
up
to
one
millimolar,
because
too
many
miles
was
a
complete
disaster.
Okay,.
A
E
A
B
A
That
that
unit,
the
aromatic
ring
with
the
sulfur
is,
is
a
decent
leaving
group,
because
the
S
minus
essentially
is
you
know
able
to
be
stabilized
by
the
ring
next
door.
So
there
are
some
medicinal
chemists
who
say
you
know.
You've
got
to
be
really
careful
with
structures
like
this,
that
they
are
promiscuous,
I,
guess
one
answer
to
that
would
be.
Is
it
the
case
that
there
are
molecules
in
this
set
which
have
that
exact
Motif
which
are
not
active.
B
I
would
say
that
these
compounds
have
actually
gone
through
two
different
sets
of
assays
and
they've
come
through,
so
there
is
some
selectivity
associated
with
them.
It
can
also
be
said
that
we
had
to
adopt
a
second
last
night
to
get
this
far,
because
we
did
have
issues
with
off
Target
activity
right.
B
D
That's
correct
so
yeah
just
some
people
might
take
off
Target
activities
like
a
toxicity
or
something
like
that.
It's
not
necessarily
toxicity.
It
was
just
interference
with
the
S,
the
reagents
and
and
binding
to
the
you
know.
Some
of
the
reagents
you're
using
in
the
assays
anyway
go
ahead.
Chris
I
mean
Matt.
A
D
Are
well
no,
no,
we
were
I.
Guess
we'll
put
it
this
way.
Maybe
just
naive.
If
you
want
to
be
harsh
about
it,
I
mean
we
didn't.
We
didn't
filter
I'm
sure
we
filtered
some.
A
D
A
D
Think
I
mean
it's
definitely
it's
definitely,
you
know
would
be
a
concern.
There.
I
mean
I,
guess
this
question
becomes.
You
know
like
what
are
there
various
whether
it's
in
we
are
like
Ace,
your
D
I
really
have
a
look
to
see
what
cysteines
are
there,
for
example,
and
then
the
other
question
would
be
obviously
all
the
reagents
that
are
in
the
assay
itself.
D
You
know
or
any
of
those
going
to
be
something
that
would
would
be
a
good
good
Target,
for
you
know
this
basically
clipping
off
the
the
amino
acid
yeah.
B
I
mean
one
thing:
I
could
do,
for
example,
the
assay
system
at
the
moment
is
insensitive
to
the
addition
of
files
which
the
amplex
red
aslay
wasn't.
So
I
could
repeat
the
answer
saying
well,
I,
don't
know,
10
millimolent
die
size
virtual
and
if
the
inhibitions
of
need
disappears,
then
that
might
be
a
clue.
Yeah
right.
A
A
Yeah,
but
also
you
know
it
in
in
the
whole,
set
of
the
of
the
molecules,
if
there's
a
way
of
clustering
by
similarity
and
finding
them,
you
know
the
molecules
that
are
most
similar
to
this
one
and
seeing
whether
they
also
are
hit.
Then
again,
it
just
gives
you
some
Faith
as
to
whether
this
is
just
a
promiscuous
class
or
whether
actually
we're
getting
binding
of
this
compound.
It
doesn't.
B
I
mean
the
other
thing
to
remember
and
I
have
to
keep
on
remodeling
myself
is
that
these
are
fragments,
and
because
of
that,
the
expectations
that
they're
going
to
have
selectivity
is
probably
a
little
bit
ambitious.
In
fact,
experience
shows
that
that's
most
certainly
the
case
and
the
selectivity
that
you
require,
and
you
want,
is
going
to
be
engineered
subsequently
based
upon
what
the
molecule
is
and
what
the
market
will
become,
because
this,
of
course,
is
the
first
applicable
work
in
progress.
A
D
D
You
know
so
this
Amino
pyrimidine
bioprimity
core
I
mean
obviously
that's.
You
can
see
that
as
part
of
kind
of
probably
in
theory,
that's
what's
supposed
to
be
mimicking
the
adenine
portion
and
then
the
thyodiazole
is
kind
of
going
into
where
we
were
hoping
like
some
of
the
phosphate
binding
spaces.
I
mean
that's
the
rough
thing,
so
I
guess
the
one
comment:
Adrian,
no
I,
I
I'm
glad
you
pointed
out
two
things:
one:
the
hill
slopes.
Clearly
those
are
red
flags,
I
guess.
D
The
only
other
thing
is
the
you
know.
What
you
also
addressed
was
the
solubility
issue,
and
so,
while
some
of
the
compounds
I
think
I
know
that
the
person
the
earlier
ones
where
you're
only
getting
like
you
know,
70
inhibition,
like
you,
said,
they're,
basically
crapping
out
as
you
get
these
higher
concentrations,
so
it
doesn't
necessarily
mean
that
they're
I
don't
know
if
I
would
necessary,
de-prioritize
them
I,
just
think
you
know
you're,
not
you
they
may
be
binding
is
just
you
can't
tell
you
know
these
higher
concentrations
and
getting
a
good
curve.
D
So
anyway,
what
it
will
be
interesting
to
see
how
they
look
and
maybe
against
the
other,
the
other
isozymes,
and
do
you
see
any
kind
of
you
know
if
you
start
seeing
at
least
some
level
of
multi-targeting,
then
maybe
that
will
at
least
maintain
some
interest
in
them.
I
guess
put
it
that
way.
D
This
is
great
to
see.
You
know,
think
clearly.
A
D
B
E
A
Mean
you
know
we
would
obviously
hope
to
try
to
take
the
most
promising
compounds
and
again
do
a
little
search
for
what's
commercially
available
yeah
around
those
compounds
if
we
picked
the
best.
But
you
know,
as
with
any
fragment
project,
The
Next
Step
might
be
to
take
the
best
and
try
and
get
structures.
Crystal
structures
of
them
bound
to
protein
is.
B
Is
that
you
know
feasible,
Laura
has
seen
this
she's
and
we
have
chatted
about
actually
progressing
these
things.
I
mean
the
other
thing
that
I'm
also
going
to
have
to
do
is
order
in
fresh
com,
fresh
samples
of
these
11.,
simply
if
it's
simply
for
just
to
be
sure
about
what
we're
looking
at
so
to
make
sure
there's
no
batch
to
batch
variability
and
so
on
and
so
forth.
D
B
Also,
we
are
running
out
of
some
of
the
compounds
in
the
enemy
library
because
it's
been
so
heavily
used,
so
we
will
be
ordering
in
fresh
compound
for
the
crystal
trials
and
also
fixed
in
the
entomology
further.
A
B
A
So
yeah
that's
a
mixture
of
the
the
Clusters
around
the
existing
hits
from
this
and
from
the
atomized
compounds
and
it's
Ed's
final
compounds
for
the
competition
and
then
it's
Yi,
Wei,
six
or
so
compounds
that
she
finished
synthesizing.
Okay,
so
yeah
that
that
reached
about
a
month
ago,
Yi
Wei's
handing
in
in
about
a
month
or
so
so
we
were
very
much
hoping
that
there
may
be
data
on
on
those
just
for
her
thesis.
A
She'll
update
you
in
a
second.
What.
D
Really
good
I
was
gonna,
I'm,
sorry,
yeah,
yeah,
yeah,
definitely
I,
just
as
as
going
forward,
I
think
one
of
the
things
Adrian
and
I
I
understand.
You
definitely
probably
should
try
to
prioritize
some
of
the
UCL
things,
especially
if
people
are
trying
to
get
their
thesis
written,
but
at
one
point
it'd
be
good
to
go
back.
D
I
think
we
need
to
test
the
AZ
compounds,
the
original
ones,
that
that
were
we
got
from
AstraZeneca
in
your
essay
because
there's
you
know
clearly
there's
I'm
going
to
call
it.
There
was
a
disconnect
right
between
what
was
published
in
terms
of
potency
because
the
different
you
know
different
assay
conditions,
so
at
least
at
some
point
having
the
relative
values
of
these
fragments
versus
in
your
assay.
You
know
what
what
are
the
AZ
compounds?
D
Actually
that
were
published
that
were
supposedly
nanomolar,
Inhibitors
I
think
you
were
seeing
almost
like
a
thousand
volt
difference,
so
at
some
point,
I
think
going
back
with
now.
You've
done
all
this
validation
of
your
assay
to
run
the
AC
compound,
so
we
have
when
we
publish,
which
we
obviously
want
to
do.
We
can
publish
you
know,
okay,
this
is
the
data
for
those
AZ
compounds
and
our
assay,
and
this
is
what
the
fragments
look
like.
So
we
can
compare
those.
B
Yeah
I
remember
this,
that
the
problem
we
had
if
I
remember
rightly,
was
that
the
enzyme
concentrations
were
using
were
quite
High,
so
we'll
be
using
sort
of
sub
micromolar
concentrations
as
in
you
know,
100
will
sell
an
animal
and
the
potency
of
these
things
was
probably
10
20
nanomolar
in
the
AZ
assays.
So
it's
something
we've
got
to
go
back
to
yeah.
D
I'm
just
saying
that's,
that's
this
at
some
point.
That's
a
I've
put
on
as
an
action
item,
but
again
it
doesn't
necessarily
have
to
happen
next
month
or
whatever
I
mean
it's
just
it's
it's
something
that,
as
we
look
forward
to
even
Grant
applications
and
Publications
I,
think
having
and
also
I
know,
Chris
said
that
he
had
gotten
inquiries
from
astronic
itself
kind
of
wondering.
You
know
what
what
had
happened
with
those
compounds
they
donated
to
us
so
yeah,
just
just
put
it
on
the
on
the
action
item
list
at
some
point,
yeah.
B
D
B
I
saw
a
paper
just
before
the
meeting
that's
been
circulated
about
adding
positive
charges
to
improve
playmeability.
It's
the
paper
that
was
just
sent
around
by
right,
I
think
there's
another
paper,
much
older
one
from
Chris
Walsh's
lab,
where
they've
chained,
with
where
they've
inserted
on
the
scenes
and
larger
means
into
the
sequence
of
an
antibiotic
or
taracidum,
which
is
a
gram-positive
targeting
molecule,
and
it
extends
sped
some
quite
nicely
Instagram
negative,
so
I'll
circulate
that
just
in
case
it's
interesting.
A
Yeah,
please
do
yeah
you,
you
hang
his.
His
current
targets
are
taking
the
AZ
compound,
so
I'm
jumping
in
for
you
hang
just
because
this
has
come
up
now
and
if.
A
Know
we
we've
sent
the
the
aiming
compound,
so
so
changing
the
oh
here
in
the
original
comment
to
this
nh2,
which
appears
to
be
e-fluxed
out
but,
but
you
hang
is,
is
still
in
need
of
a
of
a
of
an
ic50
against
the
purified
enzyme
for
that
compound
to
see
whether
it
actually
works,
but
the
next
step
there
is
to
play
around
with
these
guanidinium
and
pyridinium
groups
to
see
whether
or
not
we
can
use
some
of
that
idea
to
to
improve
things
yeah
so
yeah.
A
This
is
his
his
current
Target
list
is
this:
okay.
E
Yes,
yes,
I
think
it's
pretty
funny,
I'm
fine
with
it.
Also
we
we
have
formed
out
the
first
chapter,
seasonal
car.
A
Yeah
we'll
do
that
in
a
sec
just
one
second:
okay,
yeah,
okay
yeah,
so
this
is
I
mean
essentially
once
once.
We've
ex
we've
explored
the
oh
nh2
conversion,
the
primary
aiming
and
then
we've
explored
the
introducing
charge
groups.
Then
you
hang
has
a
complete
piece
of
work
that
we
need
to
publish,
because
we've
tried
to
we've
tried
to
improve
the
AZ
compounds
directly
and
we'll
see
if
it
works,
but
these
are
all
the
strategies
that
we
can
think
of.
I.
Think
okay,.
A
And
then,
since
I've
got
the
screen
share,
Heroes
posted
this
update
of
what
she's
been
doing
on
one
of
the
atomized
compounds.
So
you
weigh
a
few.
A
few
compounds
were
sent
already
as
part
of
the
last
shipment,
which
you've
not
shown
here,
but
that's
fine.
Do
you
just
want
to
explain
what
you've
been
trying
to
do
here.
C
Which,
hopefully,
resolve
the
solubility
problem
of
the
family?
C
A
A
Zhang's
doing
an
MSC
project
over
the
summer
and
has
been
making
some
molecules
here
that
are
intended
to
be
additions
to
the
AZ
compounds,
but
in
a
different
quadrant
so
to
in
line
with
with
Joe's
original
idea
here
of
growing
the
molecule
in
another
Direction
to
to
make
more
productive
interactions
with
the
phosphate
binding
part
of
the
of
the
protein.
A
So
so
these
subunits
here
the
the
Jack's
trying
to
make
a
part
of
that
project.
So
yeah
he's
he's
making
quick
progress
here
on
making
some
useful
groups
that
we
can
add
onto
the
molecule
quite
nitrogen,
heavy
and
and
yeah
so
eohank
did
you
want
to
just
give
a
lightning
update
on
the
cc4
car
proposal
because
obviously
we're
trying
to
get
money
to
progress
that
chemistry
a
little
bit
yeah.
D
E
So
this
is
how
we
this
is
the
this
so-called
proposal.
I'll
first
introduce
okay.
Why
we're
gonna
make
this
this
compounds
and
how
we're
going.
E
It
briefly,
then:
let's
try
the
rationale.
The
first
thing
is
like
okay,
the
idea
Behavior
series
or
like
appearing
to
have
consistent
binding
modes
with
pseudomonas
Originals
Mercy,
and
these
are
the
supporting
crystallography
data.
Sorry
Crystal
structures
that
we
did
overlay
to
illustrate
this
manner,
and
the
second
thing
is
the
second
thing:
is
the
I
there's
a
helix
appearing
there's
a
helix,
apparently
very,
pretty
consistent
sitting
in
the
position
among
different
different
murders,
from
Mercy
to
Murray
and
across
different
species.
E
Therefore
we
were
thinking.
Okay,
we
can
Target
on
this
Helix
by
growing
our
Asic
Ada
structures
into
this
pocket,
and
then,
after
this
we.
E
Yep
yeah,
after
that,
we
just
I
just
compile
the
whole
rational.
Why
we're
gonna
change
this
bits
or
we're
gonna
change?
How
we're
gonna
change
these
two
positions
in
order
to
Target
on
the
Helix,
whether
why
we're
gonna
keep
these
two
nitrogens
to
to
still
have
the
key
binding
interactions
with
the
residues
in
the
ATP
pocket.
Why
we're
gonna
change
this
part?
E
This
part,
whether
we
need
to
cut
it
off
or
we
need
to
try
different
electron
desert
distribution
and
also
why
we
need
to
cut
this
off
to
in
order
to
avoid
the
potential
clashes
with
the
Helix
from
the
central
from
the
sea
terminals.
So
that's
basically
the
rationale.
Then
we've
done
the
yeah.
So
this
is
one
of
our
examples
that
we
designed
based
on
all
these
criterias,
and
we
did
the
constraint
docking
to
illustrate
okay,
how
how
the
interactions
can
be
improved
across.
E
Oh
sorry,
there's
a
mistake
here:
there's
a
mercy
and
worthy
yeah,
sorry,
there's
an
overlay
of
mercy
and
worthy.
So
that's
kind
of
okay.
Well,
we
do
get
a
huge
Improvement
in
adding
the
hydrogen
bondings
around
this
area,
while
still
keeping
the
core
sitting
where
the
ISA
compound
was
sitting
and
having
that
key
interaction
with
the
with
this
sorry,
let
me
see
it's
a
larger.
A
D
E
Anti-Microbial
testing
is
whether
we
have
the
enough
resources
to
set
it
up
or
any
other
suggestions.
Like
any
other
suggestions
on
setting
up
the
experiments
any
other
testings,
we
need
to
involve
so
that
we
could
make
the
proposal
complete.
A
Okay,
so
just
remind
everyone
here:
this
is
a
proposal
for
chemistry
done
by
others,
so
we're
requesting
chemistry
resource,
but
the
molecules
would
need
to
be
tested
in
some
way
and
we
have
to
provide
a
suggested
way
that
we
would
do
that.
So
once
we're
happy
with
the
proposal
and
we're
going
to
need
to
talk
to
you
guys
at
Warwick
about
about
how
much
we
can
commit
to
that,
and
if
we
can't
commit
all
the
stuff
how
we
might
come
up
with
a
solution
for
testing
the
molecules
that
they
make.
A
All
right
so
I've
got
to
run
for
a
train
in
10
minutes.
I
can
I
can
hand
over
the
meeting
to
somebody
else,
but
I
wanted
to
make
sure
everyone
else
had
a
chance
to
talk.
Laura
did
you
have
anything
you
wanted
to
to
talk
about.
C
And
you
hit
yeah
and
I'll,
try
that
and
see
if
we
can
get
some
structure
with
those.
C
There's
some
already
with
like
nice
course
and
heel.
A
C
So
I'll
keep
trying
that
I'm
preparing
for
Cocoa,
crystallization
and
also
soaking
right
I'll.
Try
everything
I
can
okay.
A
Right,
wonderful,
thanks,
really
good!
Okay,
all
right
great!
Was
there
anything
anyone
else
wanted
to
raise.